Next Article in Journal
Inversion of Initial Field Based on a Temperature Transport Adjoint
Next Article in Special Issue
Assessment of Risks Associated with Extreme Climate Events in Small-Scale Bivalve Fisheries: Conceptual Maps for Decision-Making Based on a Review of Recent Studies
Previous Article in Journal
Techno-Economic Assessment of Offshore Wind Energy in the Philippines
Previous Article in Special Issue
The Effects of Temperature and Salinity Stressors on the Survival, Condition and Valve Closure of the Manila Clam, Venerupis philippinarum in a Holding Facility
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
JMSE
Volume 9
Issue 7
10.3390/jmse9070759
jmse-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Main Drivers of Fecundity Variability of Mussels along a Latitudinal Gradient: Lessons to Apply for Future Climate Change Scenarios

by
Gabriela F. Oliveira
1,*,
Hanifah Siregar
2,
Henrique Queiroga
1 and
Laura G. Peteiro
3,*
1
CESAM—Centre for Environmental and Marine Studies, Department of Biology, University of Aveiro, 3810-193 Aveiro, Portugal
2
Wildlife Conservation Society, Bogor 16128, Indonesia
3
Instituto de Investigaciones Marinas (IIM), Consejo Superior de Investigaciones Científicas (CSIC), 36208 Vigo, Spain
*
Authors to whom correspondence should be addressed.
J. Mar. Sci. Eng. 2021, 9(7), 759; https://doi.org/10.3390/jmse9070759
Submission received: 16 June 2021 / Revised: 2 July 2021 / Accepted: 5 July 2021 / Published: 9 July 2021
(This article belongs to the Special Issue Impacts of Climate Change on Shellfisheries)
Browse Figures
Versions Notes

Abstract

:
Bivalve relevance for ecosystem functioning and human food security emphasize the importance of predictions of mussel performance under different climate stressors. Here, we address the effect of a latitudinal gradient of temperature and food availability on the fecundity of the Mediterranean mussel to try to better parameterize environmental forcing over reproductive output. We show that temperature plays a major role, acting as a switching on–off mechanism for gametogenesis, while food availability has a lower influence but also modulates the number of gametes produced. Temperature and food availability also show different effects over fecundity depending on the temporal scale evaluated. Our results support the view that the gametogenesis responds non-linearly with temperature and chlorophyll concentration, an issue that is largely overlooked in growth, production and energy budgets of bivalve populations, leading to predictive models that can overestimate the capability of the mussel’s populations to deal with climate change future scenarios.

1. Introduction

The accumulation of evidence on climate change impact on marine ecosystems increasingly points towards complex cascading effects at different levels of biological organization [1,2,3,4,5]. Rising atmospheric CO2 has consequences not only for ocean temperature and acidification, but also on circulation patterns, stratification, or oxygen content [3,4]. Physical and chemical changes derived from anthropogenic activity have direct and indirect effects on physiology, behaviour and interactions of marine organisms, with consequences at population and community levels, compromising ecosystem functioning and finally its ability to provide a range of goods and services to humans [2,3,4,5]
Understanding and foreseeing the response of marine populations to climate change is crucial for human planning and development of mitigation and adaptation measures [3,6,7]. A commonly employed approach is to couple population and climate models to test for the effect of different climate scenarios on species distribution, performance or productivity [3,7,8,9]. Many of these studies agree to identify reef forming calcified organisms (corals, mussels, oysters, etc.) as one of the most vulnerable, because of their particular sensitivity to ocean acidification and other cascading effects related to the rising temperatures [1,3,10].
Mussels are amongst the most studied reef forming organisms, not just because of their interest as ecological engineers, driving biodiversity patterns [11] or regulating benthic-pelagic coupling process and water quality [12,13], but also because of their relevance for aquaculture and food security [14]. Recommendations of scientific advisory committees for a sustainable response to the increasing human demand of animal protein include an increment of bivalve production by 100 Mt before 2030 [15]. The large vulnerability of bivalves coupled to their relevance for food security emphasizes the importance of accurate predictions of mussel performance under different climate change scenarios, and has prompted the use of this species as a model for the study of the impact of several stressors [16,17,18,19].
Reproductive success is a key process controlling population dynamics, and identifying the environmental drivers that control reproductive output is key to be able to forecast the consequences of climate change [7,20,21]. The reproductive cycle of bivalves and other invertebrates comprises a sequence of biological events comprising gametogenesis, spawning and a resting period before gonad restoration [22]. This cycle is controlled by endogenous and environmental factors, and their interactions influence the onset and the duration of the reproductive cycle [22]. Among these environmental factors, temperature and food availability are considered to be the main factors regulating reproduction in marine invertebrates [20,21,22,23,24]. Temperature is usually regarded as a trigger for spawning, but also regulates the rate of gonadal development with a cumulative effect [20,21,22,23,24]. Food availability has been directly associated to fecundity (number of gametes produced) [24,25] although some authors also pointed to phytoplankton blooms as a spawning cue, which allows matching larval development with optimal environmental conditions [26].
Under unfavourable environmental conditions, bivalves activate costly defence and repair mechanisms, which reallocate energy away from reproduction towards somatic maintenance [16,20]. Therefore, reabsorption of gametes and fecundity loss are expected consequences of stressing scenarios which will become more persistent in the future [1,3]. Nonetheless, studies forecasting reproductive output and performance of bivalve populations, based on the Dynamic Energy Budget (DEB) principle, consistently predict a maintenance or a slight increase of reproductive output for bivalves in all the locations where warming future scenarios don’t overpass thermal tolerance of the species [7,9,20,23,27,28,29,30].
DEB and other models assume that physiological rates are proportional to size and follow at least a temperature-dependent relationship according to the optimal physiological range for somatic maintenance of the species, but do not usually consider specific environmental limits for gamete development [31,32,33]. Even when many attempts have been made to increase the complexity of the description of energy allocation to reproduction to make it more realistic [23], specific non-linear responses of gamete production to environmental conditions are not taken into account. Optimum values of temperature and food availability might differ between somatic and gametogenic processes. For example, while temperature tolerance ranges of Mytilus galloprovincialis range from 5 °C to 35 °C, with an optimum for somatic maintenance around 17.5 °C [28], some studies have determined a reproduction limiting temperature of 19 °C for the same species, with an optimum for the production of vitellogenic oocytes around 14 °C [34].
A correct parameterization of the complex relationships between environmental drivers and gamete production requires a better understanding of the underlying processes. Knowledge about the scales of spatial and temporal variability in reproductive output coupled to environmental variability is essential to gain a better understanding of the reproductive physiology, and to be able to incorporate it properly in the prediction of future scenarios.
In this study, we examined the variability in reproductive outputs of a series of populations of Mytilus galloprovincialis along a latitudinal gradient covering the Portuguese coast during the main reproductive season (Spring) for two different years. Consistency of latitudinal patterns of fecundity between years along with environmental variability was explored to identify relationships along an environmental gradient. Our goal was to identify non-linear relationships between gamete production and environmental variables at different temporal scales, to understand their influence when looking at the complete gametogenesis process (≈4 months) or to the last phases of gamete maturation. That information might help to provide a better parameterization of reproduction, and to develop more accurate models for mussel populations forecasting.

2. Materials and Methods

2.1. Sampling Design

Mediterranean mussels, Mytilus galloprovincialis, were collected at 7 sites along the Portuguese coastline in 2014 and 2017 (Figure 1) following a latitudinal gradient from North to South: Póvoa do Varzim, Costa Nova, Peniche, Cascais, Galapos and Zavial. Sampling locations ranged from around 37° N to 41° N (distance ≈ 560 km, mesoscale), and took place during spring in 2014 and 2017 (14 April to 4 May 2014 and 18 May to 22 June 2017). At each site, 150 individuals were collected randomly along the intertidal, for length frequency and fecundity analysis.

2.2. Environmental Variables

Sea Surface Temperature (SST), Sea Surface Salinity (SSS) and Chlorophyll-a concentration (Chl-a) as environmental variables were obtained from the beginning of the year to the sampling date at each location. Daily averaged values of SST (°C) and SSS of 2014 were provided by Hycom model (http://www.hycom.org) and values of Chl-a concentration (mg/m3) were obtained from interpolation of satellite data (MODIS, SeaWiFS and MERIS) supplied by CERSAT/IFREMER (http://ftp.ifremer.fr/ifremer/cersat/products/gridded/ocean-color/atlantic/EURL4-CHL-ATL-v01) (both accessed on 4 July 2014). The SST and SSS values of 2017 were obtained from the Mercator Ocean (Daily Global Analysis and Forecast of Ocean Physics, http://www.mercator-ocean.fr, EU Copernicus Marine Environment Monitoring Service, accessed on 7 November 2017), resolution 1/12° (≈7 km for medium latitudes). Values of Chl-a concentration (mg/m3) were obtained from CMEMS Products (Copernicus-GlobColour Project, ACRI-ST, EU Copernicus Marine Environment Monitoring Service, accessed on 7 November 2017), which consist of daily data interpolated and reprocessed (multi-year time series) from satellite observations (SeaWIFs, MODIS-Aqua, MERIS and VIIRSN), and have a horizontal resolution of 1 km. Based on these data, a set of variables regarding temperature and food availability were created to identify different effects of environmental variability, at long and short-time scales, on fecundity. Concerning short-term effects, monthly averaged SST (average SST, °C) and Chl-a concentration (average Chl-a, mg/m3) during the month preceding the sampling date were calculated for each location. Data from 2014 for Cascais, Sines and Zavial were unavailable, so these were not included in the analysis. Looking at the long-term effects, we calculated the averaged Chl-a concentration during 10 days around the peak of the spring transition (Max Chl-a) and the number of days with SST over 14 °C (T > 14 °C) during the 4 months previous to the sampling date for each location to match the expected gametogenesis duration [35,36]. The threshold of 14 °C was selected based on the temperature optimum reported in laboratory experiments for the production of vitellogenic oocytes [34].

2.3. Fecundity Analysis

All mussels collected were measured along their anterior–posterior axis to calculate their total length (L, mm), using a digital calliper. Samples were divided into size classes with 10 mm intervals, from 5 mm to 95 mm (no individuals larger than 100 mm were detected in the intertidal). From each size class, 5 females were randomly selected for fecundity analysis. Females were first identified visually based on the orange coloration of their gonads and the presence of oocytes confirmed under a stereomicroscope. From the gonads of each female, a 1 cm diameter sample was taken to assess fertility, according to the methodology presented in Sukhotin and Flyachinskaya [37]. Each gonad sample was weighed, gently macerated and diluted with 20–25 mL of salted water. From this dilution, 3 sub-samples between 150–250 µL were taken to count the oocytes under a microscope, with the help of a Bogorov counting chamber. Another gonad sample (1 cm) was taken to estimate the weight of the tissue used for oocytes quantification. The rest of the gonad was separated from the remaining soft tissues to estimate the weight of gonads with regard to the rest of tissues. After dissecting, all samples were dried at 60 °C for 24 h to obtain dry weights.
Absolute fecundity (AF, number of oocytes/female) and relative fecundity (RF, number of oocytes/g gonad) were estimated according to the following formulas:
AF =   N ¯ (   Wg Ws Vd Vs 100 95   )   RF = AF Wg
where   N ¯ is the average number of oocytes observed in 3 subsamples, Wg is the dry weight of the total gonad (g), Ws is the dry weight of the gonad sample extracted for the oocyte count (g), Vd is the volume of water where the oocytes were diluted (mL), and Vs is the volume used in the subsamples for counting oocytes (mL). Since the extraction of oocytes from the samples is only 95% efficient, with 5% of oocytes remaining in the tissues during the extraction process [37], a 100/95 correction factor was applied to the estimate of AF.

2.4. Statistical Analysis

In order to evaluate the temporal and spatial variability of the AF and RF we used factorial ANOVAs with location and year as fixed factors. The relationship between size and AF was analysed with ANCOVA (Separate Slopes model), separately for each year, including size (L, mm) as a covariable and location (Loc, 7 levels) as a fixed factor. In these analysis AF, RF and Size were log-transformed. The data and residues were evaluated for normality (Q-Q plot and Shapiro-Wilk Normality Test) and homoscedasticity (Levene’s Test). ANOVAs and ANCOVAs were performed using the software Statistica 12 (Stat Soft, Tulsa, OK, USA).
Generalized additive models (GAM), as implemented in the mgcv library of R 3.6.2 [38], were used to investigate the seasonal patterns (day of the year) and local variation (location) of the environmental variables measured (SST, SSS and Chl-a). GAM models were also employed to investigate the relationships between the environmental parameters (days > 14.5, average SST, average SSS, max Chl-a) on RF (log-transformed). The Akaike information criterion (AIC) was used to select the optimal set of variables for inclusion in the model. Model validation included the verification of homogeneity, normality and independence assumptions [39].

3. Results

3.1. Environmental Variables

Seasonality largely explained the variability observed in temperature and salinity (77.7 to 86.5% of deviance explained by the variables location and day of the year; Table 1, Figure 2). Latitudinal patterns were consistent during both sampling years, with a temperature and salinity increase from North to South (Table 1). Nonetheless, the estimated coefficients revealed warmer temperatures and lower salinities in 2017 (Table 1).
On the other hand, day of the year had a lower effect on the variability of Chl-a, with around 20%–30% of variance explained, much less than that observed for SSS or SST (Table 1; Figure 2). Both years maintained a similar latitudinal pattern for Chl-a, with a decreasing trend from North to South but with a maximum in Costa Nova (Table 1). Nonetheless, during 2014, the Chl-a peak related to spring transition was much more pronounced than during 2017 (Figure 2), and averaged estimated parameters also indicated larger concentrations of Chl-a during 2014 (Table 1).

3.2. Fecundity

Both indicators of fecundity, AF and RF, showed a significant interaction between location and year (Table 2). With regard to the absolute fecundity, larger values were observed in 2017 in comparison with 2014, and in spite of the significant interaction between year and location (Table 2), the latitudinal pattern was quite similar during both sampling years (Figure 3A). With the exception of Zavial, which showed an opposite trend between years, the larger number of eggs produced per individuals were consistently detected at Costa Nova Peniche and Cascais (Figure 3A). Average size of the mussel populations also showed a predominance of larger individuals at those locations (Figure 3B).
A positive linear relationship between AF and female size was observed for both sampling years, in all locations but Sines in 2014 (Table 3 and Table 4). In this case, the lack of relationship between both variables might be related to the narrow range of sizes sampled (25.5–35.9 mm) in that particular location and year. Although AF maintained a positive relationship with size in every location, the significant interaction between location and size indicates differences among locations on the relevance of size on determining the amount of oocytes produced (Table 3) and therefore on the value of the fitted slopes (Figure 3C; Table 4). Nonetheless, similar slopes were fitted for each location at different years (with the exception of Sines), suggesting this relationship to be quite stable at each location (Figure 3C; Table 4).
When looking at the relative fecundity values, larger RF values were reached in 2014 in comparison to 2014, but also a different spatial pattern was observed between years (Figure 4; Table 2). During 2014, RF values were low at the north and central coast locations and increased almost exponentially towards the southernmost populations (Galapos, Sines and Zavial). Conversely, during 2017, RF increased linearly from the northernmost location to the central coast, reaching a maximum at Cascais which was followed by a sharp drop in RF towards the southernmost locations (Figure 4).
According to the AIC, the four variables used to assess the effect of long-term (T > 14 °C and Max. Chl-a) and short-term (Average SST and Average Chl-a) environmental variability on RF were included in the GAM, explaining 26.6% of the variability observed across years and locations (Table 5 and Table 6; Figure 5). Despite max. Chl-a not showing a significant effect on RF, the variable was maintained in the model because of its traction over AIC (Table 5 and Table 6; Figure 5D). The number of days with temperatures over 14 °C during the previous 4 months was the variable with the larger influence on RF (lowest p-value; Table 6). The highest positive effect of T > 14 °C was around 80 days, with a sharp detrimental effect on RF when warm days overpassed 100 days (Figure 5A). Average SST and Average Chl-a during the previous month had a similar influence on RF (p-values; Table 6), but while Average Chl-a showed a positive and linear relationship with RF (Figure 5C), Average SST only seemed to exert an effect when temperatures rose above 16 °C (Figure 5B).

4. Discussion

Temperature latitudinal gradients are usually explored to understand the effects of thermal stress on marine invertebrates and to try to predict the subsequent impacts of future climate change scenarios over their distribution and population dynamics [21,40,41,42]. Our study explores the reproductive output of mussel populations along a latitudinal gradient at the North Atlantic south limit of Mytilus galloprovincialis distribution. The sampled populations cover a temperature range of 1.5 °C from the northernmost to the southernmost locations (Table 1) which encompass the predicted rise of SST by 2100 under certain climate change scenarios (1.2 to 3.6 °C for RCP2.6 and RCP8.5 respectively) [1]. The temperature latitudinal gradient observed in our study area goes along with a trend of rising salinity and decreasing Chl-a availability from North to South (Table 1). Nonetheless, while the temperature differences between extreme locations was stable between years (1.5 °C even when average temperature was much larger in 2017 than in 2014), for salinity and Chl-a, spatial variability was not so consistent (Table 1). Maximum Chl-a and minimum salinity values were detected at the northernmost location in both years, even if lower riverine inputs and a much larger phytoplankton bloom during spring transition occurred in 2014 (Figure 2), but the progressive trend towards the south was not monotonic (Table 1). Other local factors like riverine inputs, upwelling intensity or topography may have a larger influence over the spatio-temporal variability of those variables, preventing in many cases a direct correlation of Chl-a and salinity with latitude [21].
The large spatio-temporal variability on relevant environmental variables (food availability, wave intensity, air exposure, topographical characteristics, etc.) coupled to complex interactions and trade-offs between them and temperature, has been frequently suggested as an explanation for the lack of detection of latitudinal gradients on physiological performance attributable to an increase of thermal stress [21,40,41,42]. Particularly, for mussel reproductive output, the lack of detection of a thermal stress latitudinal gradient is commonly attributed to the large influence of food availability on the reproductive processes [7,20,23]. Food availability regulates the number of oocytes produced [25], but also modulates stress response capability and metabolic activity in mussels, increasing their ability to cope with different sources of stress [43]. Some authors identify more complex interactions between environmental variables, but associated to oceanographic processes which lead to the persistence of multivariate mosaics of optimal environmental conditions and therefore maximum reproductive investment [21].
Our results showed that 3 locations in the Central-North coast of Portugal (Peniche, Cascais and Costa Nova) registered the largest absolute fecundity values during both years (Figure 3A). Egg production per individual was much higher in 2017 than in 2014, suggesting a larger influence of temperature over food availability for the reproductive output of those populations, since 2017 was characterized by warmer temperatures and lower food availability (Table 1; Figure 2). Nonetheless, the higher values observed for AF at those locations, both during 2014 and 2017, seems to be driven by the larger size of mussels at these three populations (Figure 3B). Mussels, as well as many iteroparous invertebrates, show a direct relationship between reproductive effort and body size, which is usually explained as an age-related increase on the proportion of energy allocated to reproduction [24,25,44,45]. Our results highlight the relationship between size and AF for all the populations and years studied (Table 3). Nonetheless, slopes describing the relationship between AF and size varied between locations (Figure 3C), which might be a consequence of the amount of stress experienced at each population. Previous studies have suggested that environmental stress can push populations of mussels to reach reproductive maturity at smaller sizes [46,47], as well as to attain earlier their maximum rate of reproductive investment, resulting in steeper slopes between reproductive effort and size [44]. Optimal environmental conditions usually lead to less stressed and faster growing populations, which reach reproductive maturity at larger sizes but also attain larger life spans [44,45,46,47]. Several authors have reported a decay on reproductive output on mussels over 50 mm, related to senescence processes [37,44,48], which can also flatten the slope of the relationship between oocyte production and size in populations where average size was larger (Figure 3). A combination of both processes might determine the relationship of reproductive output with size at each population [44], and agree with the hypothesis of persistent multivariate mosaics of environmental conditions which determine the adaptation of each population to their local habitat [21], and comply with the consistence of the slopes calculated for each location between years in the present study (Table 4; Figure 3C).
The standardization of AF by gonad weight (RF) allows for the integration of size on the reproductive output of each population, and for the detection of further latitudinal patterns which vary between years (Figure 4). In 2014, RF increased almost exponentially towards the south, suggesting almost a linear effect of temperature on reproductive output over certain thresholds (Table 1; Figure 2). Nonetheless, in 2017 maximum RF values were displaced towards the central coast of Portugal, with a sharp drop of RF farther South (Figure 4), which seems to indicate that the warmer temperatures detected that year (Table 1; Figure 2) overpassed reproductive optimums at those locations, even when temperatures recorded were far from the upper thermal limit reported for the species (≈27–35°C; [7,28]).
Gonadal development of mussels and many other bivalves takes place over a range of temperatures which are usually well below their thermal tolerance limit for survival [34,49,50,51]. Temperature influences metabolic rates, increasing energetic cost and limiting the energy available for reproduction [24]. For M. galloprovincialis, energy allocation for reproduction seems to decrease fast when temperature rises over the optimum for somatic maintenance (≈17.5 °C; [28]), leading to a reproduction limiting temperature around 19 °C [34]. RF latitudinal gradients observed in the present study (Figure 4) might reflect the thermal history of mussels during gametogenesis according to those limits, reaching maximum values at those locations and years where temperature (Figure 2) was between the reported optimum (≈14 °C; [34]) for M. galloprovincialis gamete production and its upper limit (19 °C; [34]). Food availability can compensate for the extra demand of energy under certain stress conditions [43], but several studies report decreased rates of gametogenesis under high temperatures for bivalves, even under no limiting food conditions [34,50,51].
Our GAM model also reflects a larger effect of the temperature history over the food availability on gamete production (Table 6), supporting the role of temperature as a switching-on/-off mechanism for gametogenesis (Figure 5A). The number of days with temperatures over the optimum for gamete production during gametogenesis (4 months) explained the largest amount of the variability observed on RF (p-value < 2.9 × 10−5; Table 6). RF increases slightly with the number of warm days until those overpasses certain limit (≈80 days; Figure 5A) leading to a sharp drop on reproductive output. “Temperature windows” for gametogenesis are well documented for bivalves [25] and their evolution is usually associated with the advantage of matching reproduction with optimal conditions for larval development [52]. For M. galloprovincialis in the Iberian Coast, two peaks of reproduction are commonly reported (spring and autumn) [36,53], when food availability and oceanographic conditions optimize larval survival [52,54]. Therefore, having temperature thresholds for gametogenesis to avoid energy allocation for gamete production during winter and summer, when larval survival is limited by environmental conditions, might be a good strategy for energy optimization.
It is well known that temperature has multiple effects over bivalves’ gametogenesis, first acting as an on/off mechanism, and once activated regulating its kinetics [55]. Our results also highlighted a positive effect of temperature over RF when it rises over 16 °C during the last month of gametogenesis (Figure 5B). The average temperature values recorded during our study did not reach the reproduction limiting temperature (19 °C [34]) at any time (Figure 5B). Therefore, the positive relationship observed between temperature and RF during the last month of gametogenesis is not expected to be maintained over that temperature threshold.
We also detected a linear and positive effect of Chl-a concentration during the last month of gametogenesis over RF (Figure 5C), while the maximum Chl-a attained during the spring bloom had no significant effect on the reproductive output (Figure 5D; Table 6). These results might indicate that mussels are taking an opportunistic reproductive strategy, since the reproductive output is only dependent of the most recent food concentration. Food availability has a direct effect over the amount and quality of gametes produced by bivalves, but according to their reproductive strategy, species can be classified as opportunistic or conservative [24,56]. In opportunistic species the gametogenesis is linked to the food supply and occurs when feeding conditions are favourable, while in the conservative species the nutrients are stored before gametogenesis is initiated [24]. Mussels are, however, quite flexible and can exhibit both types of strategy, adapting it to the prevailing environmental conditions [24]. For some authors, gametogenesis initiation is mostly limited by nutrient availability, either as reserves or recently ingested food [45], while for others it is the temperature window which determines initiation/cessation of gametogenesis, and food availability modulates the amount of gametes produced [24]. Our results support better the second hypothesis, since we detected a much lower influence of food availability than temperature over RF (Table 6). The patterns observed (Figure 5) also agree with a major role of accumulative temperature conditions acting as a switching-on/off mechanism for gametogenesis, while more recent environmental conditions (temperature and food availability) regulate gamete production rates.
IPPC predicted a rise of SST by 2100 ranges from 1.2 to 3.6 °C (for RCP2.6 and RCP8.5 respectively; [1]). Our results suggest that under these climate predictions, temperature windows for gametogenesis will limit the number of reproductive events and might also alter the reproductive phenology of mussels to avoid persistent warm temperatures which will start earlier in the year and last for longer. Many studies agree on predicting changes on reproductive phenology of mussels [7,9,28,52], because an earlier activation of gametogenesis but an upper temperature limit controlling the decline of gametogenesis is not commonly included. Therefore, even when mismatch effects of dislodging reproduction towards less favourable conditions for larval survival might negatively affect mussel population dynamics [52,53], many studies based on DEB models consistently predict a maintenance or a slight increase of reproductive output for bivalves, and an increase of the population performance in all the locations where warming future scenarios don’t overpass thermal lethal tolerance of the species [7,9,20,23,27,28,29,30]. As mentioned before, those results might be related to the lack of specific parametrization of non-linear responses of gamete production to environmental conditions. DEB models establish general rules and priorities for energy allocation (somatic maintenance > growth > maturity maintenance > reproduction), and estimations of the cost of different physiological processes [31]. In addition, DEB models establish a series of assumptions for bivalve reproduction which usually include: a threshold of degree-days to activate gametogenesis, a threshold of gonadosomatic index to achieve ripe condition and a temperature threshold to trigger spawning [7,9,20,23,27,28,29,30]. However, those models do not include an upper limit for gametogenesis [9,20,23,27,28,29,30]. With the exception of some studies which incorporate a thermal upper limit for spawning [7], the lack of a specific upper limit for gametogenesis and the assumption that reproduction follows the same thermal tolerance rules than other physiological processes [31] forces the predictions towards large reproductive outputs [9,20,23,27,28,29,30], even at temperatures over the reported limit for reproduction [34]. The lack of correct parameterization of gametogenesis can therefore overestimate mussel population performance under future climate change scenarios.
Our results deepen the understanding of the underlying processes controlling the complex interactions between environmental drivers and gamete production, particularly highlighting the relevance of considering the thermal window for gametogenesis. Our study also points out the relevance of a correct parameterization of those processes, to be able to perform accurate predictions of population performance under future climate change scenarios.

5. Conclusions

Our study supports the view that gametogenesis responds non-linearly to temperature and chlorophyll concentrations, and that these variables show different effects over fecundity depending on the temporal scale evaluated. We found a major role of long-term temperature, acting as a switching on-off mechanism for gametogenesis, while short-term food availability has a lower influence but also modulates the amount of gametes produced.
Our results highlight the complexity of gametogenesis environmental dependences, and how its inclusion in predictive models is vital to avoid overestimations on the capability of mussel populations to deal with climate change future scenarios.

Author Contributions

Conceptualization, H.Q. and L.G.P.; methodology, G.F.O., H.S., H.Q., L.G.P.; formal analysis, G.F.O., H.S., H.Q. and L.G.P.; investigation, G.F.O., H.S., H.Q. and L.G.P.; resources, H.Q.; data curation, G.F.O., H.S., H.Q. and L.G.P.; writing—original draft preparation, G.F.O. and L.G.P.; writing—review and editing, G.F.O., H.Q. and L.G.P.; supervision, H.Q. and L.G.P.; funding acquisition, G.F.O., H.Q. and L.G.P. All authors have read and agreed to the published version of the manuscript.

Funding

This study is part of the ‘Larval Sources-Assessing the ecological performance of marine protected area networks’ research project, funded by Fundação para a Ciência e Tecnologia-FCT/MCTES (PTDC/BIA-BIC/120483/2010). Financial support was allocated under the COMPETE Programme, which includes components from the European Regional Development Fund and from the Ministério da Ciência, Tecnologia e Ensino Superior. Thanks are also due to FCT/MCTES for the financial support to CESAM (UIDP/50017/2020+UIDB/50017/2020), through national funds. G.F.O. was supported by a grant from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior–Brasil (CAPES)–Finance Code 001, grant number 13724/13-4 L.G.P. was supported by a postdoctoral fellowship from Xunta de Galicia, Spain (POS-B/2016/032) during the field work and by a Talento Senior Grant (16_IN585A_2020_986937) during the writing of the manuscript. H.S. was supported by a grant from the Erasms Mundus Master of Science in Marine Biodiversity and Conservation.

Institutional Review Board Statement

Ethical review and approval were waived for this study because this is a very abundant and non-endangered species. Samples taken were only those strictly necessary for the analyses, and were made under permission of the Agência Portuguesa do Ambiente and the Instituto de Conservação da Biodiversidade e Florestas.

Informed Consent Statement

Not applicable.

Data Availability Statement

The datasets generated and analyzed during this study are available from the corresponding author upon request.

Acknowledgments

The authors would like to thank Inês Gomes, Rui Albuquerque and Rafael Batista (Universidade de Aveiro), as well as Cláudia Moreira and Ana Margarida Gama (Centro Interdisciplinar de Investigação Marinha e Ambiental, Universidade do Porto) and João Castro (Laboratório de Ciências do Mar da Universidade de Évora) for their support during the fieldwork and laboratory analysis. We also would like to thank Maria de Jesus Fernandes, Director of the Departamento de Conservação da Natureza e Florestas de Lisboa e Vale do Tejo of the Instituto de Conservação da Natureza e Florestas, who issued sampling permits and provided access to facilities of the Reserva Natural das Berlengas (RNB) and the Parque Natural da Arrábida (PNA). Licenses for rocky shore sampling outside the protected areas were provided by the Agência Nacional do Ambiente.

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

  1. Bindoff, N.L.; Cheung, W.W.L.; Kairo, J.G.; Arístegui, J.; Guinder, V.A.; Hallberg, R.; Hilmi, N.; Jiao, N.; O’Donoghue, S.; Suga, T.; et al. Changing Ocean, Marine Ecosystems, and Dependent Communities. In IPCC Special Report on the Ocean and Cryosphere in a Changing Climate; Intergovernmental Panel on Climate Change: Geneva, Switzerland, 2019; pp. 477–587. [Google Scholar]
  2. Doney, S.C.; Ruckelshaus, M.; Duffy, J.E.; Barry, J.P.; Chan, F.; English, C.A.; Galindo, H.M.; Grebmeier, J.; Hollowed, A.B.; Knowlton, N.; et al. Climate Change Impacts on Marine Ecosystems. Annu. Rev. Mar. Sci. 2012, 4, 11–37. [Google Scholar] [CrossRef] [Green Version]
  3. Gattuso, J.-P.; Magnan, A.K.; Billé, R.; Cheung, W.W.L.; Howes, E.L.; Joos, F.; Allemand, D.; Bopp, L.; Cooley, S.R.; Eakin, C.M.; et al. Contrasting futures for ocean and society from different anthropogenic CO2 emissions scenarios. Science 2015, 349, aac4722. [Google Scholar] [CrossRef]
  4. Hoegh-Guldberg, O.; Bruno, J.F. The Impact of Climate Change on the World’s Marine Ecosystems. Science 2010, 328, 1523–1528. [Google Scholar] [CrossRef]
  5. Scheffers, B.R.; De Meester, L.; Bridge, T.C.L.; Hoffmann, A.; Pandolfi, J.; Corlett, R.; Butchart, S.H.M.; Pearce-Kelly, P.; Kovacs, K.M.; Dudgeon, D.; et al. The broad footprint of climate change from genes to biomes to people. Science 2016, 354, aaf7671. [Google Scholar] [CrossRef]
  6. Santos, C.F.; Agardy, T.; Andrade, F.; Calado, H.; Crowder, L.B.; Ehler, C.N.; García-Morales, S.; Gissi, E.; Halpern, B.S.; Orbach, M.K.; et al. Integrating climate change in ocean planning. Nat. Sustain. 2020, 3, 505–516. [Google Scholar] [CrossRef]
  7. Thomas, Y.; Bacher, C. Assessing the sensitivity of bivalve populations to global warming using an individual-based modelling approach. Glob. Chang. Biol. 2018, 24, 4581–4597. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  8. Froehlich, H.E.; Gentry, R.R.; Halpern, B.S. Global change in marine aquaculture production potential under climate change. Nat. Ecol. Evol. 2018, 2, 1745–1750. [Google Scholar] [CrossRef] [PubMed]
  9. Steeves, L.E.; Filgueira, R.; Guyondet, T.; Chassé, J.; Comeau, L. Past, Present, and Future: Performance of Two Bivalve Species Under Changing Environmental Conditions. Front. Mar. Sci. 2018, 5, 184. [Google Scholar] [CrossRef]
  10. Hare, J.A.; Morrison, W.E.; Nelson, M.W.; Stachura, M.M.; Teeters, E.J.; Griffis, R.; Alexander, M.; Scott, J.D.; Alade, L.; Bell, R.J.; et al. A Vulnerability Assessment of Fish and Invertebrates to Climate Change on the Northeast U.S. Continental Shelf. PLoS ONE 2016, 11, e0146756. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  11. Smith, J.R.; Fong, P.; Ambrose, R.F. Dramatic declines in mussel bed community diversity: Response to climate change? Ecology 2006, 87, 1153–1161. [Google Scholar] [CrossRef]
  12. Froján, M.; Figueiras, F.G.; Zúñiga, D.; Pérez, F.A.; Arbones, B.; Castro, C.G. Influence of Mussel Culture on the Vertical Export of Phytoplankton Carbon in a Coastal Upwelling Embayment (Ría de Vigo, NW Iberia). Estuaries Coasts 2016, 39, 1449–1462. [Google Scholar] [CrossRef] [Green Version]
  13. Petersen, J.K.; Saurel, C.; Nielsen, P.; Timmermann, K. The use of shellfish for eutrophication control. Aquac. Int. 2015, 24, 857–878. [Google Scholar] [CrossRef]
  14. Food and Agriculture Organization (FAO). The State of World Fisheries and Aquaculture 2020: Sustainability in Action; The State of World Fisheries and Aquaculture (SOFIA); Food and Agriculture Organization: Rome, Italy, 2020. [Google Scholar]
  15. Science Advice for Policy by European Academies. Food from the Oceans: How Can More Food and Biomass Be Obtained from the Oceans in a Way That Does Not Deprive Future Generations of Their Benefits; SAPEA: Berlin, Germany, 2017. [Google Scholar]
  16. Farcy, É.; Burgeot, T.; Haberkorn, H.; Auffret, M.; Lagadic, L.; Allenou, J.-P.; Budzinski, H.; Mazzella, N.; Pete, R.; Heydorff, M.; et al. An integrated environmental approach to investigate biomarker fluctuations in the blue mussel Mytilus edulis L. in the Vilaine estuary, France. Environ. Sci. Pollut. Res. 2012, 20, 630–650. [Google Scholar] [CrossRef] [PubMed]
  17. Fly, E.K.; Hilbish, T.J.; Wethey, D.; Rognstad, R.L. Physiology and Biogeography: The Response of European Mussels (Mytilus spp.) to Climate Change. Am. Malacol. Bull. 2015, 33, 136–149. [Google Scholar] [CrossRef] [Green Version]
  18. Kroeker, K.J.; Gaylord, B.; Hill, T.M.; Hosfelt, J.D.; Miller, S.H.; Sanford, E. The Role of Temperature in Determining Species’ Vulnerability to Ocean Acidification: A Case Study Using Mytilus galloprovincialis. PLoS ONE 2014, 9, e100353. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  19. Zippay, M.L.; Helmuth, B. Effects of temperature change on mussel, Mytilus. Integr. Zoöl. 2012, 7, 312–327. [Google Scholar] [CrossRef]
  20. Gourault, M.; Petton, S.; Thomas, Y.; Pecquerie, L.; Marques, G.; Cassou, C.; Fleury, E.; Paulet, Y.-M.; Pouvreau, S. Modeling reproductive traits of an invasive bivalve species under contrasting climate scenarios from 1960 to 2100. J. Sea Res. 2019, 143, 128–139. [Google Scholar] [CrossRef] [Green Version]
  21. McCabe, M.; Navarrete, S. Reproductive investment in rocky intertidal mussels: Spatiotemporal variability and environmental determinants. Mar. Ecol. Prog. Ser. 2018, 599, 107–124. [Google Scholar] [CrossRef]
  22. Gosling, E. Reproduction, Settlement and Recruitment. In Bivalve Molluscs: Biology, Ecology and Culture; Fishing News Books: Oxford, UK, 2003; pp. 131–168. [Google Scholar]
  23. Bernard, I.; de Kermoysan, G.; Pouvreau, S. Effect of phytoplankton and temperature on the reproduction of the Pacific oyster Crassostrea gigas: Investigation through DEB theory. J. Sea Res. 2011, 66, 349–360. [Google Scholar] [CrossRef] [Green Version]
  24. Seed, R.; Suchanek, T. Population and community ecology of Mytilus. In the Mussel Mytilus: Ecology, Physiology, Genetics and Culture; Gosling, E., Ed.; Elsevier: San Diego, CA, USA, 1992; pp. 87–170. [Google Scholar]
  25. Gosling, E. (Ed.) Reproduction, Settlement and Recruitment. In Marine Bivalve Molluscs; John Wiley & Sons: Oxford, UK, 2015; pp. 157–202. [Google Scholar]
  26. Starr, M.; Himmelman, J.H.; Therriault, J.-C. Direct Coupling of Marine Invertebrate Spawning with Phytoplankton Blooms. Science 1990, 247, 1071–1074. [Google Scholar] [CrossRef]
  27. Monaco, C.J.; McQuaid, C. Climate warming reduces the reproductive advantage of a globally invasive intertidal mussel. Biol. Invasions 2019, 21, 2503–2516. [Google Scholar] [CrossRef]
  28. Montalto, V.; Helmuth, B.; Ruti, P.M.; Dell’Aquila, A.; Rinaldi, A.; Sarà, G. A mechanistic approach reveals non linear effects of climate warming on mussels throughout the Mediterranean sea. Clim. Chang. 2016, 139, 293–306. [Google Scholar] [CrossRef]
  29. Sara, G.; Palmeri, V.; Rinaldi, A.C.; Montalto, V.; Helmuth, B. Predicting biological invasions in marine habitats through eco-physiological mechanistic models: A case study with the bivalve Brachidontes pharaonis. Divers. Distrib. 2013, 19, 1235–1247. [Google Scholar] [CrossRef] [Green Version]
  30. Sara, G.; Palmeri, V.; Montalto, V.; Rinaldi, A.; Widdows, J. Parameterisation of bivalve functional traits for mechanistic eco-physiological dynamic energy budget (DEB) models. Mar. Ecol. Prog. Ser. 2013, 480, 99–117. [Google Scholar] [CrossRef]
  31. Kooijman, B. Dynamic Energy Budget Theory for Metabolic Organisation; Cambridge University Press: Cambridge, UK, 2010. [Google Scholar]
  32. Pörtner, H.-O.; Gutt, J. Impacts of Climate Variability and Change on (Marine) Animals: Physiological Underpinnings and Evolutionary Consequences. Integr. Comp. Biol. 2016, 56, 31–44. [Google Scholar] [CrossRef] [Green Version]
  33. Christensen, E.A.F.; Norin, T.; Tabak, I.; van Deurs, M.; Behrens, J.W. Effects of temperature on physiological performance and behavioral thermoregulation in an invasive fish, the round goby. J. Exp. Biol. 2021, 224, 237669. [Google Scholar] [CrossRef]
  34. Fearman, J.; Moltschaniwskyj, N. Warmer temperatures reduce rates of gametogenesis in temperate mussels, Mytilus galloprovincialis. Aquaculture 2010, 305, 20–25. [Google Scholar] [CrossRef]
  35. Cáceres-Martínez, J.; Figueras, A. Long-term survey on wild and cultured mussels (Mytilus galloprovincialis Lmk) reproductive cycles in the Ria de Vigo (NW Spain). Aquaculture 1998, 162, 141–156. [Google Scholar] [CrossRef] [Green Version]
  36. Villalba, A. Gametogenic cycle of cultured mussel, Mytilus galloprovincialis, in the bays of Galicia (N.W. Spain). Aquaculture 1995, 130, 269–277. [Google Scholar] [CrossRef]
  37. Sukhotin, A.A.; Flyachinskaya, L.P. Aging reduces reproductive success in mussels Mytilus edulis. Mech. Ageing Dev. 2009, 130, 754–761. [Google Scholar] [CrossRef] [PubMed]
  38. R Core Team. R: A Language and Environment for Statistical Computing; R Foundation for Statistical Computing: Vienna, Austria, 2018. [Google Scholar]
  39. Zuur, A.; Ieno, E.N.; Walker, N.; Saveliev, A.A.; Smith, G.M. Mixed Effects Models and Extensions in Ecology with R; Springer: New York, NY, USA, 2009. [Google Scholar]
  40. Helmuth, B.; Harley, C.; Halpin, P.M.; O’Donnell, M.; Hofmann, G.E.; Blanchette, C.A. Climate Change and Latitudinal Patterns of Intertidal Thermal Stress. Science 2002, 298, 1015–1017. [Google Scholar] [CrossRef] [Green Version]
  41. Sara’, G.; Kearney, M.; Helmuth, B. Combining heat-transfer and energy budget models to predict thermal stress in Mediterranean intertidal mussels. Chem. Ecol. 2011, 27, 135–145. [Google Scholar] [CrossRef]
  42. Newell, R.I.E.; Hilbish, T.J.; Koehn, R.K.; Newell, C.J. Temporal Variation in the Reproductive Cycle of Mytilus edulisl. (Bivalvia, Mytilidae) from Localities on the East Coast of the United States. Biol. Bull. 1982, 162, 299–310. [Google Scholar] [CrossRef]
  43. Dowd, W.W.; Felton, C.A.; Heymann, H.M.; Kost, L.E.; Somero, G.N. Food availability, more than body temperature, drives correlated shifts in ATP-generating and antioxidant enzyme capacities in a population of intertidal mussels (Mytilus californianus). J. Exp. Mar. Biol. Ecol. 2013, 449, 171–185. [Google Scholar] [CrossRef]
  44. Bayne, B.L.; Salkeld, P.N.; Worrall, C.M. Reproductive effort and value in different populations of the marine mussel, Mytilus edulis L. Oecologia 1983, 59, 18–26. [Google Scholar] [CrossRef]
  45. Bayne, B.; Worrall, C. Growth and Production of Mussels Mytilus edulis from Two Populations. Mar. Ecol. Prog. Ser. 1980, 3, 317–328. [Google Scholar] [CrossRef]
  46. Sprung, M. Reproduction and fecundity of the mussel Mytilus edulis at helgoland (North sea). Helgol. Mar. Res. 1983, 36, 243–255. [Google Scholar] [CrossRef] [Green Version]
  47. Seed, R. The ecology of Mytilus edulis L. (Lamellibranchiata) on exposed rocky shores. Oecologia 1969, 3, 277–316. [Google Scholar] [CrossRef] [PubMed]
  48. Rodhouse, P.; McDonald, J.H.; Newell, R.I.E.; Koehn, R.K. Gamete production, somatic growth and multiple-locus enzyme heterozygosity in Mytilus edulis. Mar. Biol. 1986, 90, 209–214. [Google Scholar] [CrossRef]
  49. Beukema, J.J.; Dekker, R.; Jansen, J.M. Some like it cold: Populations of the tellinid bivalve Macoma balthica (L.) suffer in various ways from a warming climate. Mar. Ecol. Prog. Ser. 2009, 384, 135–145. [Google Scholar] [CrossRef]
  50. Heasman, M.; O’Connor, W.; Frazer, A. Temperature and nutrition as factors in conditioning broodstock of the commercial scallop Pecten fumatus Reeve. Aquaculture 1996, 143, 75–90. [Google Scholar] [CrossRef]
  51. Martínez, G.; Pérez, H. Effect of different temperature regimes on reproductive conditioning in the scallop Argopecten purpuratus. Aquaculture 2003, 228, 153–167. [Google Scholar] [CrossRef]
  52. Philippart, C.J.M.; Van Bleijswijk, J.D.L.; Kromkamp, J.C.; Zuur, A.F.; Herman, P. Reproductive phenology of coastal marine bivalves in a seasonal environment. J. Plankton Res. 2014, 36, 1512–1527. [Google Scholar] [CrossRef] [Green Version]
  53. Philippart, C.J.; Amaral, A.; Asmus, R.; Van Bleijswijk, J.; Bremner, J.; Buchholz, F.; Cabanellas-Reboredo, M.; Catarino, D.; Cattrijsse, A.; Charles, F.; et al. Spatial synchronies in the seasonal occurrence of larvae of oysters (Crassostrea gigas) and mussels (Mytilus edulis/galloprovincialis) in European coastal waters. Estuar. Coast. Shelf Sci. 2012, 108, 52–63. [Google Scholar] [CrossRef] [Green Version]
  54. Peteiro, L.G.; Labarta, U.; Fernández-Reiriz, M.; Alvarez-Salgado, X.; Filgueira, R.; Piedracoba, S. Influence of intermittent-upwelling on Mytilus galloprovincialis settlement patterns in the Ría de Ares-Betanzos. Mar. Ecol. Prog. Ser. 2011, 443, 111–127. [Google Scholar] [CrossRef] [Green Version]
  55. Navarro, E.; Iglesias, J.I.P. Energetics of reproduction related to environmental variability in bivalve molluscs. Haliotis 1995, 24, 43–55. [Google Scholar]
  56. Bayne, B.L. Aspects of reproduction in bivalve molluscs. In Estuarine Processes; Wiley, M., Ed.; Academic Press: New York, NY, USA, 1976; pp. 432–448. [Google Scholar]
Jmse 09 00759 g001 550
Figure 1. Fecundity sampling locations along the Portuguese coastline.
Figure 1. Fecundity sampling locations along the Portuguese coastline.
Jmse 09 00759 g001
Jmse 09 00759 g002 550
Figure 2. Estimated smoothing curves of day of the year for Sea Surface Temperature (SST, °C), Sea Surface Salinity (SSS) and Chlorophyll-a concentration (Chl-a, mg/m3) from 2014 and 2017, expressed as a mean between locations along the Portuguese coastline. Results of the General Additive Model showing the effect of the variable “day of the year” (Julian year). Dashed lines show a 95% Confidence Interval and tick marks along the x-axis represent when observations occurred. Vertical dashed line (2017) is a reference to the last day from 2014 (day 120).
Figure 2. Estimated smoothing curves of day of the year for Sea Surface Temperature (SST, °C), Sea Surface Salinity (SSS) and Chlorophyll-a concentration (Chl-a, mg/m3) from 2014 and 2017, expressed as a mean between locations along the Portuguese coastline. Results of the General Additive Model showing the effect of the variable “day of the year” (Julian year). Dashed lines show a 95% Confidence Interval and tick marks along the x-axis represent when observations occurred. Vertical dashed line (2017) is a reference to the last day from 2014 (day 120).
Jmse 09 00759 g002
Jmse 09 00759 g003 550
Figure 3. Average and standard error of (A) absolute fecundity (AF, n°. oocytes per individual), (B) Size (mm) and (C) Coefficients Slopes of Analysis of Covariance (ANCOVA) between AF (logarithm-transformed) and Size (log), significant values (p < 0.05), except to Sines. Data of the different locations along the Portuguese coastline, from North (left) to South (right), during the years 2014 and 2017.
Figure 3. Average and standard error of (A) absolute fecundity (AF, n°. oocytes per individual), (B) Size (mm) and (C) Coefficients Slopes of Analysis of Covariance (ANCOVA) between AF (logarithm-transformed) and Size (log), significant values (p < 0.05), except to Sines. Data of the different locations along the Portuguese coastline, from North (left) to South (right), during the years 2014 and 2017.
Jmse 09 00759 g003
Jmse 09 00759 g004 550
Figure 4. Average and standard error of relative fecundity (RF, n°. oocytes per gram of gonad) in the different locations along the Portuguese coastline, from North (left) to South (right), during the years of 2014 and 2017.
Figure 4. Average and standard error of relative fecundity (RF, n°. oocytes per gram of gonad) in the different locations along the Portuguese coastline, from North (left) to South (right), during the years of 2014 and 2017.
Jmse 09 00759 g004
Jmse 09 00759 g005 550
Figure 5. Results of the General Additive Model showing the best subset of environmental variables explaining relative fecundity (RF, log) along the Portuguese coastline, during 2014 and 2017 (grouped years). (A) Number of days with Sea Surface Temperature higher than 14 °C at last 4 months (N Days SST > 14 °C). (B) Average of Sea Surface Temperature (Av. SST, °C) and (C) Chlorophyll-a (Av. Chl-a, mg/m3) at last month before samples. (D) Chl-a concentration during 10 days around the peak of the spring transition (max. Chl-a). Dashed lines show 95% Confidence Interval and tick marks along the x-axis represent where the RF values (points) occurred.
Figure 5. Results of the General Additive Model showing the best subset of environmental variables explaining relative fecundity (RF, log) along the Portuguese coastline, during 2014 and 2017 (grouped years). (A) Number of days with Sea Surface Temperature higher than 14 °C at last 4 months (N Days SST > 14 °C). (B) Average of Sea Surface Temperature (Av. SST, °C) and (C) Chlorophyll-a (Av. Chl-a, mg/m3) at last month before samples. (D) Chl-a concentration during 10 days around the peak of the spring transition (max. Chl-a). Dashed lines show 95% Confidence Interval and tick marks along the x-axis represent where the RF values (points) occurred.
Jmse 09 00759 g005
Table
Table 1. Structure of the General Additive Model describing Sea Surface Temperature (°C), Sea Surface Salinity (SSS) and Chlorophyll-a concentration (mg/m3) variability along the coastline from January to May 2014 and January to June 2017. S.E.: standard error; e.d.f.: estimated degrees of freedom.
Table 1. Structure of the General Additive Model describing Sea Surface Temperature (°C), Sea Surface Salinity (SSS) and Chlorophyll-a concentration (mg/m3) variability along the coastline from January to May 2014 and January to June 2017. S.E.: standard error; e.d.f.: estimated degrees of freedom.
SST (°C)SSSChl-a (mg/m3)
2014Parametric CoefficientsParametric CoefficientsParametric Coefficients
LocationEstimateS.E.tpEstimateS.E.tpEstimateS.E.tp
Póvoa do Varzim
(Intercept)		13.3570.031436.581p < 0.00135.2940.0172051.909p < 0.0012.4990.13821.488p < 0.001
Costa Nova0.2870.044−6.541p < 0.0010.1030.024−4.239p < 0.0010.4760.196−6.427p < 0.050
Peniche0.4860.04411.077p < 0.0010.6110.02425.197p < 0.001−1.2580.196−6.427p < 0.001
Cascais0.6670.05212.883p < 0.0010.4180.02914.651p < 0.001----
Galapos0.7300.04416.614p < 0.0010.5120.02421.103p < 0.001−0.6980.196−3.563p < 0.001
Sines1.1030.05221.310p < 0.0010.6110.02924.423p < 0.001----
Zavial1.5710.06225.430p < 0.0010.7460.03421.903p < 0.001----
Smooth terms (non parametric)Smooth terms (non parametric)Smooth terms (non parametric)
e.d.fFpe.d.fFpe.d.fFp
Day of the Year8.94302.6p < 0.0016.0734.04p < 0.0018.6116.3p < 0.001
R2 adjusted: 0.861% Deviance explained: 86.5%R2 adjusted: 0.772% Deviance explained: 77.7%R2 adjusted: 0.278% Deviance explained: 29.6%
2017Parametric coefficientsParametric coefficientsParametric coefficients
LocationEstimateS.E.tpEstimateS.E.tpEstimateS.E.tp
Póvoa do Varzim
(Intercept)		15.0510.057276.091p < 0.00133.2150.0321059.370p < 0.0011.7180.08024.265p < 0.001
Costa Nova0.6230.080−7.757p < 0.0011.1120.046−24.270p < 0.0010.2240.113−0.1980.843
Peniche0.1740.0802.168p < 0.0500.6670.04614.550p < 0.001−0.4960.113−4.384p < 0.001
Cascais0.5690.0807.084p < 0.0010.9400.04620.520p < 0.001−0.1800.113−1.5930.111
Galapos0.9210.08011.477p < 0.0011.4040.04630.640p < 0.001−0.4950.113−4.373p < 0.001
Sines1.1490.08014.306p < 0.0011.5620.04634.090p < 0.001−1.1320.113−10.004p < 0.050
Zavial1.4890.08018.544p < 0.0011.7590.04638.380p < 0.001−1.2410.113−10.970p < 0.001
Smooth terms (non parametric)Smooth terms (non parametric)Smooth terms (non parametric)
e.d.fFpe.d.fFpe.d.fFp
Day of the Year8.94767.6p < 0.0018.88119.5p < 0.0018.51812.03p < 0.001
R2 adjusted: 0.862% Deviance explained: 86.3%R2 adjusted: 0.846% Deviance explained: 84.8%R2 adjusted: 0.208% Deviance explained: 21.7%
Table
Table 2. Results of Factorial Analysis of Variance (ANOVA) of absolute fecundity (AF, log) and relative fecundity (RF, log) with locations and year as fixed factors (d.f.: estimated degrees of freedom, MS: mean squares).
Table 2. Results of Factorial Analysis of Variance (ANOVA) of absolute fecundity (AF, log) and relative fecundity (RF, log) with locations and year as fixed factors (d.f.: estimated degrees of freedom, MS: mean squares).
AF (log)
EffectdfMSFp
Location63.9189.950<0.001
Year115.47239.280<0.001
Loc x Year62.1735.520<0.001
Error2550.394--
RF (log)
EffectdfMSFp
Location61.4908.030<0.001
Year10.1080.5800.446
Loc x Year62.59013.960<0.001
Error2550.186--
Table
Table 3. Results of Analysis of Covariance (ANCOVA) with Separate Slopes of Absolute Fecundity (AF, log) in different locations with size (mm, log) as covariate (d.f.: estimated degrees of freedom, SS: sum of squares, MS: mean squares).
Table 3. Results of Analysis of Covariance (ANCOVA) with Separate Slopes of Absolute Fecundity (AF, log) in different locations with size (mm, log) as covariate (d.f.: estimated degrees of freedom, SS: sum of squares, MS: mean squares).
ANCOVA with Separate Slopes: AF (log) vs. Location and Size (log)
2014
EffectDfSSMSFp
Intercept11.2121.2125.197p < 0.05
Location × log size717.1522.45010.506p < 0.05
Location61.5900.2651.1370.346
Error11025.6530.233--
Total12359.150---
2017
EffectDfSSMSFp
Intercept12.3432.34310.624p < 0.05
Location × log size728.7364.10518.613p < 0.05
Location64.0150.6693.034p < 0.05
Error13128.8920.221--
Total14481.505---
Table
Table 4. Coefficients of ANCOVA with Separate Slopes analysis of absolute fecundity (AF, log) in different locations with size (mm, log) as covariate (d.f.: estimated degrees of freedom, SS: sum of squares, MS: mean squares).
Table 4. Coefficients of ANCOVA with Separate Slopes analysis of absolute fecundity (AF, log) in different locations with size (mm, log) as covariate (d.f.: estimated degrees of freedom, SS: sum of squares, MS: mean squares).
Separate Slopes Coefficients: AF (log) vs. Location and Size (log)
2014
LocationSlopeS.E.tp
Povoa do Varzim2.2680.8712.603p < 0.05
Costa Nova2.9960.7593.949p < 0.05
Peniche3.5651.2702.807p < 0.05
Cascais5.5031.2844.286p < 0.05
Galapos3.8621.0343.734p < 0.05
Sines0.6872.7370.2510.802
Zavial5.2091.5763.305p < 0.05
2017
LocationSlopeS.E.tp
Povoa do Varzim3.2371.2202.653p < 0.05
Costa Nova1.8480.8292.229p < 0.05
Peniche2.8651.0392.757p < 0.05
Cascais5.6690.7977.117p < 0.05
Galapos2.9830.5265.667p < 0.05
Sines7.7921.7534.445p < 0.05
Zavial2.6120.9152.855p < 0.05
Table
Table 5. Summary of Akaike Information Criteria (AIC) for the selection of optimal set of variables for inclusion in the model to explaining relative fecundity (RF) along the Portuguese coastline during 2014 and 2017 (grouped years). Short-term variables: average Sea Surface Temperature (Av. SST, °C) and Average Chlorophyll-a concentration (Av. Chl-a, mg/m3), both at last month before the sampling. Long-term variable: number of days with Sea Surface Temperature over 14 °C (T > 14 °C) during the 4 months before the sampling date and Average Chl-a concentration during 10 days around the peak of the spring transition (Max. Chl-a). Variables values used for the analysis were average for each location and year (d.f.: estimated degrees of freedom).
Table 5. Summary of Akaike Information Criteria (AIC) for the selection of optimal set of variables for inclusion in the model to explaining relative fecundity (RF) along the Portuguese coastline during 2014 and 2017 (grouped years). Short-term variables: average Sea Surface Temperature (Av. SST, °C) and Average Chlorophyll-a concentration (Av. Chl-a, mg/m3), both at last month before the sampling. Long-term variable: number of days with Sea Surface Temperature over 14 °C (T > 14 °C) during the 4 months before the sampling date and Average Chl-a concentration during 10 days around the peak of the spring transition (Max. Chl-a). Variables values used for the analysis were average for each location and year (d.f.: estimated degrees of freedom).
ModeldfAIC
Av. SST3.000403.786
AV. Chl-a4.799329.071
Days T > 14 °C4.947370.316
Max. Chl-a3.000353.635
Av. SST + Av. Chl-a6.656283.465
Av. SST + Days T > 14 °C5.840330.819
Av. SST + Max Chl-a6.700329.336
Av. Chl-a + Days T > 14 °C7.042286.272
Av. Chl-a + Max. Chl-a7.940296.542
Max. Chl-a + Days T > 14 °C4.950370.316
Av. SST + Av. Chl-a + Days T > 14 °C7.460284.044
Av. SST + Av. Chl-a + Max. Chl-a9.883278.559
Days T > 14 °C + Av. Chl-a + Max. Chl-a10.789278.181
Av. SST+ Days T > 14 °C + Av. Chl-a + Max. Chl-a9.943277.323
Table
Table 6. Structure of the General Additive Model describing the best subset of environmental variables explaining relative fecundity (RF) along the Portuguese coastline from 2014 and 2017, grouped. Averages of Chlorophyll-a (Av. Chl-a, mg/m3) and Sea Surface Temperature (Av. SST, °C) at last month before samples. Number of days with Sea Surface Temperature higher than 14 °C at last 4 months (N Days SST > 14 °C). Average Chl-a concentration during 10 days around the peak of the spring transition (Max. Chl-a). S.E.: standard error; e.d.f.: estimated degrees of freedom.
Table 6. Structure of the General Additive Model describing the best subset of environmental variables explaining relative fecundity (RF) along the Portuguese coastline from 2014 and 2017, grouped. Averages of Chlorophyll-a (Av. Chl-a, mg/m3) and Sea Surface Temperature (Av. SST, °C) at last month before samples. Number of days with Sea Surface Temperature higher than 14 °C at last 4 months (N Days SST > 14 °C). Average Chl-a concentration during 10 days around the peak of the spring transition (Max. Chl-a). S.E.: standard error; e.d.f.: estimated degrees of freedom.
Parametric Coefficients
EstimateS.E.Tp
Intercept5.4940.028198.300< 2 × 10−16
Smooth terms (non parametric)
e.d.fFp
N Days T > 14 °C3.08.2772.9 × 10−5
Av. SST2.47.1500.0018
Av. Chl-a1.08.1050.0048
Max. Chl-a1.60.5300.4880
R2 adjusted: 0.241% Deviance explained: 26.6%
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Oliveira, G.F.; Siregar, H.; Queiroga, H.; Peteiro, L.G. Main Drivers of Fecundity Variability of Mussels along a Latitudinal Gradient: Lessons to Apply for Future Climate Change Scenarios. J. Mar. Sci. Eng. 2021, 9, 759. https://doi.org/10.3390/jmse9070759

AMA Style

Oliveira GF, Siregar H, Queiroga H, Peteiro LG. Main Drivers of Fecundity Variability of Mussels along a Latitudinal Gradient: Lessons to Apply for Future Climate Change Scenarios. Journal of Marine Science and Engineering. 2021; 9(7):759. https://doi.org/10.3390/jmse9070759

Chicago/Turabian Style

Oliveira, Gabriela F., Hanifah Siregar, Henrique Queiroga, and Laura G. Peteiro. 2021. "Main Drivers of Fecundity Variability of Mussels along a Latitudinal Gradient: Lessons to Apply for Future Climate Change Scenarios" Journal of Marine Science and Engineering 9, no. 7: 759. https://doi.org/10.3390/jmse9070759

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop