W&M ScholarWorks
Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects
2020
Effects Of Acidification And Salinity On Callinectes Sapidus, Mercenaria Mercenaria, And Their Interactions
Katherine Sara Longmire William & Mary - Virginia Institute of Marine Science, [email protected]
Follow this and additional works at: https://scholarworks.wm.edu/etd
Part of the Ecology and Evolutionary Biology Commons
Recommended Citation Longmire, Katherine Sara, "Effects Of Acidification And Salinity On Callinectes Sapidus, Mercenaria Mercenaria, And Their Interactions" (2020). Dissertations, Theses, and Masters Projects. Paper 1616444437. http://dx.doi.org/10.25773/v5-xc1q-jy95
This Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Dissertations, Theses, and Masters Projects by an authorized administrator of W&M ScholarWorks. For more information, please contact [email protected]. Effects of acidification and salinity on Callinectes sapidus, Mercenaria mercenaria, and their interactions
A Thesis
Presented to
The Faculty of the School of Marine Science
The College of William & Mary
In Partial Fulfillment
of the Requirements for the Degree of
Master of Science
by
Katherine Sara Longmire
January 2021
APPROVAL PAGE
This thesis is submitted in partial fulfillment of
the requirements for the degree of
Master of Science
Katherine S. Longmire
Approved by the Committee, December 2020
Rochelle D. Seitz, Ph.D. Committee Chair / Advisor
Romuald N. Lipcius, Ph.D.
Iris Anderson, Ph.D.
Richard Brill, Ph.D.
Emily Rivest, Ph.D.
ii
TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS ...... v
LIST OF TABLES ...... vi
LIST OF FIGURES ...... viii
ABSTRACT ...... ix
1. INTRODUCTION ...... 2
1.1 Carbonate system chemistry in the open ocean ...... 2
1.2 Estuarine acidification…………………………………………………………3
1.3 Multiple stressors: responses to acidification and salinity………………...4
1.4 Food-web interactions and stressors………………………………………..5
1.5 Study organisms……………………………………………………………….5
1.5.1 Blue crabs …………..…………………………………………….6
1.5.2 Hard clams …………………………………………………..……8
1.6 Research rationale ……………………………………………………………8
1.7 Objectives and hypotheses …………………………………………………..9
2. METHODS …………………………………………………………………………….10
2.1 Treatment exposure………………………………………………………….10
2.2 Single-species impacts ……………………………………………………...11
2.2.1 Crab carapace strength and claw pinch force …………….....11
2.2.2 Statistical analyses: carapace strength and claw pinch force
…………………………………………………………………….11
2.2.3 Clam shell structure and ridge analysis ..…………………….12
2.2.4 Statistical analyses: clam survival, growth, and ridge rugosity
……………………………………………………………..……..12
2.3 Filmed predator-prey interactions (2019) ………………………………....13
iii
2.3.1 Statistical analyses: filmed predator-prey interactions………13
3. RESULTS ………………………………………………….………………………….14
3.1 Seawater chemistry ………………………………..………………………..14
3.2 Single-species impacts………………………………………………………15
3.2.1 Crab carapace strength ..………………………………………15
3.2.2 Crab claw pinch force ...………………………………………..16
3.2.3 Clam growth ..…………………………………………………...17
3.2.4 Clam survival…………………………………………………….18
3.2.5 Clam shell structure and ridge rugosity (2018)………………18
3.3 Filmed predator-prey interactions…………………………………………..19
4. DISCUSSION………………………………………………………………………….20
4.1 Crab carapace strength and pinch force ……………………………….…20
4.2 Clam growth, survival, shell structure, and ridge rugosity …………….....22
4.3 Predator-prey interactions ..…………………………………………………23
5. SUMMARY ..…………………………………………………………………………..24
LITERATURE CITED ...... 26
TABLES AND FIGURES ...... 35
iv
ACKNOWLEDGEMENTS
I would like to thank my advisor, Rochelle Seitz for her support and commitment, as well my committee Romuald Lipcius, Emily Rivest, Iris Anderson and Richard Brill. I am grateful for their guidance from the initial project idea to the finished thesis.
Thank you to the technicians of the Marine Conservation Ecology and Community Ecology Labs. I would not have been able to put my project into motion without their knowledge of all things blue crab and R software. Special thanks to Mike Seebo for his expertise in designing and building complex tank systems, without whom I could not have done my project.
Thank you to Jeff Shields for his help measuring crab claw pinch force.
Thank you to Olga Trofimova and Amy Wilkerson at the Applied Research Center Core Lab for training me on the tabletop SEM Phenom ProX (Nanoscience Instruments) and the digital optical microscope HIROX RH-2000 which I used to take SEM images of my clam shells.
I also would not have been able to do much without the helping hands of summer REU interns, Governor’s School interns, volunteers, and fellow students, Challen Hyman, Shantelle Landry, and Kristen Bachand.
Thank you to Hank Jones, a local clam grower, and Cherrystone Aquaculture for providing me with clams in abundance.
I would like to thank my support network at VIMS, including friends and mentors. Special thanks to Cassandra Glaspie for introducing me to the field of ocean acidification and sparking a life-long interest. Special thanks also to Jenny Dreyer for her advice on both academic and personal matters. She lifted me up and reminded me to make my own definition of success. Thank you also to Libby Jewett for her professional mentorship, showing me the inner workings of NOAA OAP and scheming with me on fun OA-related activities.
Of course, my project would not have been possible without my funding sources. Thank you to the Virginia Sea Grant, the VIMS Office of Academic Studies, the Gates Millennium Scholarship, and the National Science Foundation.
Finally, I would not have succeeded in graduate school without my family. Thank you to my mom for raising me to the woman I am today and my grandparents for showing me the wonderful world beneath ocean waves. Special thanks to my husband for his love, patience, and care every minute of every day and my cats, Sushi and Estrella, for their purrs and cuddles, and for forcefully reminding me to take breaks from my computer every so often.
v
LIST OF TABLES
1. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab carapace strength (2018) ...... 35
2. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab claw pinch force (2018) ...... 36
3. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab carapace strength (2019) ...... 37
4. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab claw pinch force (2019) ...... 38
5. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning clam growth and survival (2018 and 2019) ...... 39
6. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning clam ridge rugosity (2018) ...... 40
7. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab carapace strength (2018) ...... 41
8. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for blue crab carapace strength (2018) ...... 42
9. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab carapace strength (2019) ...... 43
10. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for blue crab carapace strength (2019) ...... 44
11. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab claw pinch force (2018) ...... 45
12. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for blue crab claw pinch force (2018) ...... 46
13. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab claw pinch force (2019) ...... 47
14. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for blue crab claw pinch force (2019) ...... 48
15. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam growth (2018) ...... 49
16. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for clam growth (2018) ...... 50
17. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam growth (2019) ...... 51 vi
18. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for clam growth (2019) ...... 52
19. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam survival (2018) ...... 53
20. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for clam survival (2018) ...... 54
21. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam survival (2019) ...... 55
22. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for clam survival (2019) ...... 56
23. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam ridge rugosity (2018) ...... 57
24. Estimate, SE, and 95% CI of the parameters from multiple linear regression models for clam ridge rugosity (2018) ...... 58
vii
LIST OF FIGURES
1. 2018 Water quality measurements over time in experimental treatments ...... 59
2. 2018 Water chemistry measurements over time in experimental treatments ...... 60
3. 2019 Water quality measurements over time in experimental treatments ...... 61
4. 2019 Water chemistry measurements over time in experimental treatments ...... 62
5. 2018 blue crab carapace strength in Newtons (N) by size in differing pH treatments ...... 63
6. 2018 blue crab carapace strength in Newtons (N) by exposure time in differing pH treatments ...... 64
7. 2019 mean (± SE) adjusted carapace strength by treatment ...... 65
8. 2018 blue crab claw pinch force in Newtons (N) by carapace width ...... 66
9. 2019 blue crab claw pinch force in Newtons (N) by carapace width ...... 67
10. Box-and-whisker plot of 2019 blue crab claw pinch force in Newtons (N) by sex…………...... 68
11. Mean slopes of 2018 and 2019 clam growth ...... 69
12. 2018 and 2019 clam survival by treatment ...... 70
13. Sample SEM images of the exterior of clam shells from different treatments ...... 71
14. Mean clam ridge rugosity (± SE) and damage index scores by treatment ...... 72
15. Stacked bar graphs of crab behavior in percentages by salinity and pH ...... 73
viii
ABSTRACT
Ocean acidification (OA) coupled with other stressors, will be detrimental at the species and ecosystem levels. Decreased carbonate ion concentrations negatively impact calcifying species, yet the combined effects of OA and other stressors are less well known, and many studies disregard species interactions. Multi-species studies involving OA and other stressors are crucial to comprehend the full threat of OA. Understanding how OA interacts with other stressors to affect species responses is necessary for future management of exploited species in an altered ecosystem. The objectives of my study were to assess: 1) the effects of long-term exposure to decreased pH and salinity on juvenile blue crab (Callinectes sapidus) carapace strength and pinch force and juvenile hard clam (Mercenaria mercenaria) armor and growth; and 2) the blue crab and hard clam predator-prey interactions under low pH and salinity via filmed mesocosm trials. In 2018 and 2019, I held juvenile blue crab (n = 24–40; 50–80 mm carapace width) and juvenile hard clams (n = 112; 10–15 mm shell length) in mesocosms with crossed pH and salinity treatments for 10 – 11 weeks. I regularly monitored water quality and chemistry and measured clams weekly. After treatment exposure, I assessed crab carapace and claw strength using force meters and imaged clam shells using a scanning electron microscope (SEM) for shell structure and ridge rugosity. In 2018, crab carapace strength increased with size, decreased with treatment exposure time, and was overall stronger after exposure to low pH. In 2019, crab carapace strength was weakest in the low-pH, high-salinity treatment and strongest in the low-pH, low-salinity treatment. In high-pH treatments, carapace strength was similar regardless of salinity. Claw pinch force also increased with size but was weaker in low-pH, in both 2018 and 2019. Moreover, male crabs in 2019 had stronger pinch force than females. Clam growth was negatively impacted by low pH in both years; however, salinity had a more negative impact in 2019 than in 2018. Clam survival followed a similar trend in both years, with survival being lowest in the low-pH, low-salinity treatment and highest in the control treatment. Deterioration of clam shell structure and ridge rugosity (indicative of damage) were also correlated, as both were greatest in the low-pH, low-salinity treatment and lowest in the control treatment. Finally, in 2019, filmed predator-prey mesocosm trials assessed the interaction between blue crabs and hard clams after treatment exposure. Low sample sizes precluded statistical analyses, but trends indicated there was no alteration in the predator-prey relationship under multiple stressors.
ix
Effects of acidification and salinity on Callinectes sapidus, Mercenaria mercenaria, and their interactions
1. INTRODUCTION
Since the Industrial Revolution, the combustion of fossil fuels has increased CO2 emissions into the atmosphere by an order of magnitude (Doney & Schimel 2007). This atmospheric CO2 has been tempered by oceanic uptake, which accounts for almost 1/3 of anthropogenic CO2 added to the atmosphere (Sabine et al. 2004, Sabine & Feely 2007).
Consequences of oceanic uptake of CO2 include alterations to the ocean carbonate system, causing reduced pH (termed “ocean acidification”, OA) (Caldeira & Wickett 2003). In the past few decades, ocean acidification has prompted numerous studies on its biological impacts
(Raven et al. 2005, Fabricius et al. 2011, Kroeker et al. 2013). Most studies have focused on the effects on a single species (Bibby et al. 2007, Landes & Zimmer 2012). While these are valuable, they are not fully representative of the complex interactions between biotic and abiotic factors. Further, recent comparisons of open ocean, coastal, and estuarine systems revealed that estuaries will react differently than the open ocean, prompting the term “estuarine acidification” (EA) (MdOA 2015). EA presents a new stress to organisms living in a highly variable environment, and combined stressors could alter the estuarine community (further described in the section on estuarine acidification below).
1.1 Carbonate system chemistry in the open ocean
Seawater carbonate chemistry is governed by a series of chemical reactions:
+ − + 2− CO2(atmos) ↔ CO2(aq) + H2O ↔ H2CO3 ↔ H + HCO3 ↔ 2H + CO3
Atmospheric CO2 combines with water to form carbonic acid (H2CO3), which then dissociates to
− 2− + form bicarbonate (HCO3 ), carbonate (CO3 ) ions, and hydrogen ions (H ). Adding CO2 increases
− + bicarbonate and hydrogen ion concentrations ([HCO3 ] and [H ], respectively), the latter of which lowers pH since pH = -log[H+]. A pH decrease of 1 unit corresponds to a 10-fold increase in
+ + 2− [H ]. Increasing [H ], however, decreases carbonate ion concentration [CO3 ], a compound necessary for the formation of calcium carbonate. The formation of calcium carbonate is very
2 important to marine organisms that rely upon it to build their shells. The concentration of atmospheric CO2 gas, pCO2, has increased dramatically in the past decades to over 406 ppm in
2017 (Keeling curve; Tans & Keeling 2017), while open ocean pH has decreased by 0.1 units since the beginning of the Industrial Revolution (Jay et al. 2018).
The carbonate alkalinity (CA) of a system represents the ability of seawater to resist pH change upon addition of an acid and is positively correlated with salinity. CA is described by the following equation:
− 2− − + CA = [HCO3 ] + 2[CO3 ] + [OH ] − [H ]
1.2 Estuarine acidification (EA)
While OA with respect to open ocean systems has now been well studied, the extent to which OA challenges estuarine communities remains an open question. Estuaries are more susceptible to acidification, as they have increased acidic inputs from terrestrial runoff and a reduced buffering capacity due to the lower salinity relative to open ocean systems (Cai & Wang
1998, Cai et al. 2017). Thus, estuaries may experience more drastic acidification sooner than the open ocean (Waldbusser et al. 2011), though the variability in estuaries makes it difficult to tease out the anthropogenic signal (Najjar et al. 2010). EA is characterized by substantial, rapid and cyclic changes in pH (MdOA 2015). Estuaries are influenced on spatial and temporal scales much more dramatically than the open ocean, and future climate scenarios predict increased variability with increased CO2 uptake (Miller et al. 2009). Besides CO2 variability, estuaries experience regular shifts in nutrient concentrations, salinity, temperature, and dissolved oxygen
(Pearl 2006). Future salinity trends in the Chesapeake Bay are particularly uncertain since the number of extreme rainfall events has been increasing over time (Jay et al. 2018), but sea level rise threatens to inundate estuaries with full-strength seawater (Hilton et al., 2008, Najjar et al.
2010, Carter et al. 2014). The pH variability and little information on the effects of multiple stressors (Denman et al. 2011, Baumann & Smith 2017) hampers our ability to predict results of stressors on food webs in coastal systems.
3
1.3 Multiple stressors: responses to acidification and salinity
2- Decreased availability of CO3 will likely impact bivalves’ shell-building abilities (Gazeau et al. 2007), but current research suggests variation among species and life stages. Larvae and juveniles are most sensitive to reduced pH (Van Colen et al. 2012). Decreased calcification
2- rates under increased CO2 levels (resulting in decreased availability of CO3 ) effect snails
(Bibby et al. 2007, Ellis et al. 2009), juvenile hard clams (Waldbusser et al. 2010), juvenile and adult oysters (Crassostrea virginica) (Miller et al. 2009, Waldbusser et al. 2011), and mussels
(Mytilus edulis) (Gazeau et al. 2007). For bivalves, shell structure is negatively impacted by simultaneous increases in CO2 and decreases in salinity (Dickinson et al. 2013); live weight and shell length are reduced by increases in CO2 alone (Bressan et al. 2014). On a macro scale, climate change has decreased the production of economically valuable mollusks (Doney et al.
2009), with a predicted loss of $100 billion by 2100 (Narita et al. 2012).
Many juvenile and adult crustaceans are more tolerant of OA physiologically in the short- term due to their ability to regulate internal pH (Whiteley 2011). Brachyuran crabs including
Callinectes sapidus (Glandon & Miller 2017) and Cancer magister (Pane & Barry 2007) have effective acid-base regulation mechanisms, allowing them to counteract short-term increases in
CO2. Under these conditions, there is either positive (Ries et al. 2009) or no significant decrease in net calcification (Glandon & Miller 2017). The ability to regulate pH is developed only as individuals age; larvae are relatively intolerant of acidification. More specifically, low pH reduces blue crab larval survival (Glitz & Taylor 2017, Tomasetti et al. 2018) and growth (Glitz & Taylor
2017) during short-term experiments.
Research regarding how blue crabs will fare during long-term exposure to increased CO2 when coupled with other stressors is, however, lacking (Whiteley 2011). For example, salinity stress coupled with other environmental factors, such as increased pCO2, could additively impact blue crabs in negative ways, but the effect of these multiple stressors remains unknown. blue crabs are strong osmoregulators and they tolerate a wide range in salinities, from near
4 freshwater to 100% seawater. Yet, low salinity causes an increased oxygen consumption, as crabs must use more energy to maintain homeostasis (King 1965, Findley et al. 1978). Fresh water can also negatively impact crab growth post molt (Tagatz 1968).
1.4 Food-web interactions and stressors
Studies on multiple stressors (involving OA) on predator-prey dynamics are needed because studies ignoring these interactions likely underestimate multi-stressor impacts (Parker et al. 2013). More specifically, OA and salinity could impact interactions among species (e.g., between bivalves and predatory crabs). Salinity plays a key role in determining community structure (Najjar et al. 2010). The effects of OA on crabs are uncertain, however, and “consumer stress theory” predicts that predators will be more affected by stress than prey (Menge &
Sutherland 1987). Combined effects of low pH and other stressors can have additive or interactive effects (Metzger et al. 2007, Miller & Waldbusser 2016), as decreased salinity exacerbates the effects of low pH (Dickinson et al. 2012).
Predation is a key determinant of the abundance and size structure of prey populations, and the structure and functioning of communities (Menge 1995, Bruno & O’Conner 2005).
However, changes in physical and behavioral reception or response to cues with environmental stress can affect predator-prey interactions. Few studies have focused on food-web interactions under lowered pH (Rosa & Seibel 2008, Amaral et al. 2012, Parker et al. 2013, Glaspie et al.
2017). Blue crabs are tactile feeders that cause sediment disturbance when foraging (Glaspie et al. 2017). When crabs are stressed under acidification, they reduce feeding rates (Dodd 2012) and increase prey handling time (Dodd et al. 2015). Acidification can also alter the number of prey consumed in mesocosm studies and the responsiveness of prey to predator cues (Glaspie et al. 2017). One mechanism used by clams to avoid crab predation is cessation of pumping activity, and this response was diminished under low pH (Glaspie et al. 2017). Time for shell hardening is also increased after molting (Lane et al. 2013). Most studies of OA and predator- prey interactions have been short-term laboratory projects (Bibby et al. 2007, Landes & Zimmer
5
2012, Amaral et al. 2012), and further tests employing long-term studies are thus warranted.
1.5 Study organisms
1.5.1 Blue crabs
The blue crab, Callinectes sapidus, is a large Portunid crab abundant in the Chesapeake
Bay (Hines et al. 1987, 1990, Lipcius & Van Engel 1990), which preys heavily upon bivalves
(Lipcius et al. 2007). Feeding efficiency (the number of prey caught over time) of blue crabs varies with prey availability, predator density, and habitat complexity (Mansour & Lipcius 1991,
Micheli 1997). Blue crabs are the main predators of adult hard clams in the Bay (Hines et al.
1990), but their diet also includes other bivalves and juvenile blue crabs, along with fish and invertebrates (Seitz et al. 2011). Blue crabs also support one of the most important fisheries in
Chesapeake Bay, with over 55 million pounds landed in VA and MD in 2019, which is worth $81 x 106 to the economy (NOAA 2019).
The carapace of blue crabs is composed of four layers: the epicuticle, exocuticle, endocuticle, and the membranous layer. The epicuticle is the outermost and thinnest layer, consisting of tanned lipoprotein and calcium carbonate. This layer may shield internal layers from the surrounding seawater, specifically in water supersaturated with CO2. The epicuticle could be the reason blue crabs have increased net calcification in elevated CO2 conditions, as this organic layer has total coverage and exposure to surrounding seawater, but the mechanism for increased net calcification under elevated CO2 conditions needs further exploration (Ries et al. 2009). The second outermost layer is the exocuticle, composed of tanned chitin-protein fibers and calcium carbonate, and hardened by quinone tanning and calcification (Travis
1955a). The third layer, the endocuticle, is the thickest and most heavily calcified and is comprised of untanned chitin-protein fibers and calcium carbonate (Travis 1955a). The innermost membranous layer is comprised of chitin and protein, but not calcium.
Blue crabs achieve growth incrementally by molting (shedding their hard carapace). Molt frequency varies with age, food availability, and temperature. Immediately upon molting, a crab
6 is extremely soft and generally inactive. Shortly after, tissue and endocuticle begin to form. The next stage, the intermolt period, is when the exoskeleton, membranous layer, epicuticle, exocuticle, and spines are completed. Then, the carapace is at its hardest and the crab is most active. Main tissue growth also occurs along with accumulation of organic reserves. Just before molting, a clear space begins to form along the ventral rim of the swim paddle as the cuticle separates from connective tissue. Finally, molting occurs when there is rapid water uptake and exuviation (Johnson 1980).
One ecological role of a blue crab’s pincer chela (i.e.. claw), is to capture prey for consumption and to crush shells. The base of each claw is a curved propodus with a movable dactyl attached so that the appendages can grab prey. The claws are heteromorphic, meaning one claw is larger than the other. The larger claw is called the “crusher”, and in most Portunid crabs is usually the right claw. The smaller claw is called the “cutter” (Hamilton et al. 1976). The crusher claw has a larger muscle volume and can generate more force (Blundon & Kennedy
1982).
Osmoregulation primarily occurs in the gills of blue crab, which consist of eight lamellae per mm of gill length (Aldridge & Cameron 1982, Taylor & Taylor 1992) and contain the sodium- potassium pump (an enzyme responsible for gill ion transport). When environmental salinity decreases to a level where the osmolality of the seawater is below that of hemolymph, the activity of sodium-potassium pump increases (Towle et al. 1976, Neufeld et al. 1980). The pump creates net Na+ uptake then transfers the Na+ from the intracellular space to the hemolymph.
Under extreme hypoosmotic conditions, increased activity of the sodium pump can strain the supply of ATP (Towle et al. 1976, Neufeld et al. 1980).
Blue crabs are also strict acid-base regulators, as osmoregulation and acid-base regulation share the same mechanisms (Whiteley et al. 2001). Adjustments of pH in the hemolymph are driven by carbonic anhydrase, an enzyme located in the gills (Burnett 1984,
Burnett & McMahon 1985, Henry 1987). Carbon dioxide is transported into the gills where it is
7 converted to bicarbonate ions by carbonic anhydrase. Strong acid-base regulators generally maintain higher bicarbonate ion concentration (Portner et al. 2004, Melzner et al. 2009).
However, most studies on crustacean acid-base regulation have been short term (Whiteley et al. 2018) and the effects of medium- to long-term elevated CO2 levels are unclear.
1.5.2 Hard clams
The hard clam, Mercenaria mercenaria, is a shallow-dwelling infaunal organism. It has short siphons that are retractable and can be sealed tightly, indicative of armor defense (Vermeij
1987), and it occupies coastal tidal environments in soft sediments throughout the lower
Chesapeake Bay (Mann et al. 2005). In Chesapeake Bay, the highest densities of hard clams are in the lower James River, where there is an important fishery for hard clams (Mann et al.
2005).
Hard clams are osmoconformers and weak acid-base regulators, such that the external environment dictates their internal chemistry (DuPaul & Webb 1974, Melzner et al. 2009). When basal maintenance is high, growth rate and tissue mass can suffer and mortality can increase
(Dickinson et al. 2013). Moreover, low salinity and acidification alter shell structure, leading to significant flaking of the periostracum layer, relative to high-salinity and high-pH conditions
(Dickinson et al. 2013).
Hard clam shells are equivalve and a consistent ovoid shape with new material added concentrically along the valve margins and across the interior surface. Each shell is composed of three aragonitic layers: an outer composite prismatic layer, a middle complex crossed- lamellar layer, and an inner complex crossed-lamellar layer. The inner and middle layers are formed by the inner mantle (Crenshaw 1980) and the outer shell layer is deposited by portions of the mantle. Clams record seasonal and temporal growth within the microstructure of the middle and outer layers.
1.6 Research rationale
8
Blue crabs and hard clams are important to Chesapeake Bay fisheries and food webs, and they provide essential ecosystem services. The depletion of a single species could upset an entire ecosystem, potentially leading to the loss of ecosystem functions and services (Duarte
2000). Bay bivalves provide habitat and can improve overall water quality, services which indirectly benefit society. The Bay also provides most US blue crab landings, a fishery that is one of the largest crustacean fisheries in the world (Miller et al. 2005), and hard clams are fished heavily in the Bay (Mann et al. 2005). These industries, which directly and indirectly affect millions of people, are predicted to be impacted by OA (Narita et al. 2012), leading to job and revenue losses. Accurate predictions of the effects of OA and other stressors on bivalves and their predators are necessary, especially for food web models (Lipcius & Latour 2006), since ignoring predator-prey interactions likely underestimates the effects of multiple stressors (Parker et al. 2013). Data on the effects of multiple stressors on crabs and clams could allow managers to make the sound decisions regarding management of resources.
1.7 Objectives and hypotheses
The objectives of my research are to:
1. assess the effects of long-term exposure to decreased pH and salinity on juvenile blue
crab carapace strength and pinch force and hard clam armor and growth, and
2. Assess predator-prey interactions under low pH and low salinity via filmed mesocosm
trials with C. sapidus and M. mercenaria
I hypothesized that, whereas salinity would not impact carapace thickness or pinch force in blue crab (given the euryhaline tolerances of blue crabs; Glandon et al. 2018), low pH would result in increased carapace thickness (as increased calcification under low pH was documented previously by Ries et al. 2009). Moreover, time of exposure to low pH would negatively impact carapace strength, but positively impact pinch force due to muscle growth during the intermolt period. I defined exposure time (ET) as the number of days a crab’s carapace was exposed to a treatment. Regarding clams, I hypothesized that concurrent low pH
9 and low salinity would negatively impact clam growth and shell structure (Dickinson et al. 2013).
Finally, I hypothesized that during predator-prey trials, clams will be easier to open and consume in low pH, regardless of salinity, but crabs will be more stressed under low pH and therefore eat fewer clams (Glaspie et al. 2017).
2. METHODS
2.1 Treatment Exposure
In 2018, I collected juvenile blue crabs (N = 24; 50 – 90 mm CW) from lower
Chesapeake Bay tributaries and juvenile hard clams (N = 112; 10 – 15 mm shell length) from a local grower on the Eastern Shore of Virginia. Animals were grown together in replicate 71-cm- diameter cylindrical tanks (3 individual crabs and 14 clams per tank within enclosures) with crossed high (30 ppt) or low (16 ppt) salinity and acidified (pH 7.0 – 7.2) or ambient (pH 8.0 –
8.3) conditions and 5-cm sand lining tank bottoms. I chose the low-pH values to represent predicted values for 2100 (Donohue et al. 2012, Glandon & Miller 2017, Glaspie et al. 2017). An automated system controlled the pH of the acidified tanks by controlling the flow CO2 being bubbled into the tanks. I manually checked temperature and salinity every other day using a YSI datasonde, and used an Omega handheld pH probe to verify the pH values measured by the automated pH controllers. The experiment ran for 11 weeks. I took water samples three times over the course of the experiment (weeks 1, 2, and 11) and analyzed each sample for dissolved inorganic carbon (DIC) using a DIC Analyzer (Model AS-C3, Apollo SciTech, LLC). Carbonate alkalinity was then calculated using the MATLAB program (using pH and DIC measurements and water quality parameters), CO2SYS (MATLAB ver. 2020a).
In 2019, juvenile blue crabs (N = 40; 45 – 72 mm CW) collected from the same lower
Chesapeake Bay tributaries as in 2018 were kept in 96-cm-diameter cylindrical tanks, whereas clams (N = 112; 10 – 15 mm shell length) were kept in rectangular tanks of 43 x 33 cm dimensions with 14 clams per tank and 5-cm sand in every tank. Tanks of similar treatments (2
10 per treatment) were connected so water could recirculate and thus minimize pH and salinity differences between the tanks. I monitored pH and water quality regularly and I took water samples for DIC in weeks 8 and 10. The experiment ran for 10 weeks in 2019.
2.2 Single-species impacts
2.2.1 Crab carapace strength and claw pinch force
Following 10 – 11 weeks of treatment exposure (crossed pH and salinity), I measured crab carapace strength and claw pinch force. Crab carapace strength was measured using a durometer, which quantifies the pressure (in Newtons) needed to deform a given object (Foyle et al. 1989). Claw pinch force was measured using custom designed instrument that quantified the force produced by a crab’s claw (Jeffrey Shields & Richard Brill, VIMS, personal communication). This claw force translates to the force a crab could use to crush a clam.
2.2.2 Statistical analyses of carapace strength and claw pinch force
I examined statistical differences of carapace strength and claw force using an
Information Theoretic approach (Burnham & Anderson 2002, Anderson 2008). I formed statistical models (gi) based on multiple alternative hypotheses regarding the influence of independent variables (pH, salinity, crab size, exposure time [defined below]) on each of the response variables (Chamberlin 1890). Models were then compared using Akaike’s Information
Criteria (AIC) with R (R Core Team 2017) and RStudio (RStudio Team 2017) statistical software
(see below for details of statistical analyses). The response variables, carapace strength and claw pinch force, were continuous. The independent continuous variables were crab size (mm
CW) and exposure time (days), defined as the number of days a crab’s carapace was exposed to a treatment since placement in the treatment or since its last molt. The factors were salinity
(low or high) and pH (low or high).
For 2018, carapace strength and claw pinch force, the global model did not involve three-way interactions:
푦 = 훽0 + 훽1푥1 + 훽2푥2 + 훽3푥3 + 훽4푥4 + 훽5푥1푥3 + 훽6푥1푥4 + 훽7푥2푥3 + 훽8푥2푥4 + 훽9푥3푥4 (1)
11 where y = carapace strength or claw pinch force, 훽0 = parameter (constant) for the baseline condition, and βi = parameters representing increases or decreases in strength because of corresponding independent variables xi. Note that the two-way interaction terms are represented as ixixj. Low sample size for both carapace strength and claw pinch force meant that this global model could not be run, thus, reduced versions of the global model were analyzed (Tables 1 and 2).
For 2019, carapace strength and claw pinch force, sample size was large enough to run a global model without three-way interactions:
푦 = 훽0 + 훽1푥1 + 훽2푥2 + 훽3푥3 + 훽4푥4 + 훽5푥5 + 훽6푥3푥4 + 훽7푥3푥5 + 훽8푥4푥5 (2) where the variables represented were the same as in 2018 (Tables 3 and 4).
2.2.3 Clam shell structure and ridge analysis
Following the 11-week treatment exposure in 2018, surviving clams were frozen for later analysis. In fall 2019, I used Scanning Electron Microscopy (SEM) to examine differences in exterior clam shell structure by treatment. Clam meat and shell were separated, dried at 60 C for 7 days, then the shell surface was imaged using an SEM. I used these images to quantify mean shell surface deterioration by treatment using the damage index scale 0 (no damage) – 3
(high damage) (Bressan et al. 2014).
Images were further used to quantify shell ridge rugosity (indicative of shell deterioration;
Dickinson et al. 2013). I traced the exterior growth ridge of a designated spot on each clam using ImageJ software and compared the length of this line to the length of an undamaged ridge line estimated on the same individual. Ridge rugosities were averaged by treatment, with the greater rugosity indicating greater damage.
2.2.4 Statistical analyses of clam survival, growth, and ridge rugosity
To analyze clam growth and shell structure, a similar logical framework as that used for crab carapace strength and claw pinch force was developed. Clam size (initial shell length),
12 salinity, and pH were independent variables similar to the crab analysis. Exposure time was not used, however, since clams do not shed their outer layer; thus, their exposure time was all the same. Clam survival and growth shared a global model:
푦 = 훽0 + 훽1푥1 + 훽2푥2 + 훽3푥3 + 훽4푥1푥2 + 훽5푥1푥3 + 훽6푥2푥3 (3) where y = clam survival or growth, 훽0 = parameter (constant) for the baseline condition, and βi
= parameters representing increases or decreases because of corresponding independent variables xi (Table 5). Note that interaction terms are represented as ixixj.
Statistical models for clam ridge rugosity involved only salinity, pH, and their interaction, included in the global model:
푦 = 훽0 + 훽1푥1 + 훽2푥2 + 훽3푥1푥2 (4) where y = clam shell ridge rugosity, 훽0 = parameter (constant) for the baseline condition, and βi
= parameters representing increases or decreases because of corresponding independent variables xi (Table 6). Note that interaction terms are represented as ixixj. Finally, I performed statistical analyses of differences among treatments, again using AIC model comparisons with
RStudio programming software.
2.3 Filmed predator-prey interactions (2019)
I determined the interactive effects of pH and salinity on predator-prey interactions via filmed mesocosm trials. Juvenile clams (10 – 12 mm shell height; susceptible to crab predation;
Eggleston et al. 1992) and crabs (50 – 80 mm carapace width) were pre-treated for 10 weeks with crossed treatments of low and high salinity and low and high pH (as in single-species exposure). Subsequently, I transplanted clams into a 0.25 m2 plot within a circular mesocosm of
96 cm diameter with 15 cm of sand at natural densities of 28 clams m-2 (Peterson et al. 1983) and 35 cm of water overlying the sand. Sample size was dependent on clam survival following treatment exposure. The low-pH, high-salinity treatment had n = 6, both low-salinity treatments had n = 2, and the control treatment had n = 4. Following pre-treatment, I starved each blue
13 crab for 24 h, then introduced it into a randomly selected replicate mesocosm with bivalves under the same pre-treatment as the crab. Crabs foraged on bivalves over a 24 h filmed predation trial (as in previous studies; Glaspie et al. 2017). For these trials, I quantified crab behavior (i.e., time spent moving, buried, and eating) and predation on clams (clam mortality).
2.3.1 Statistical analyses of filmed predator-prey mesocosm interactions (2019)
Unfortunately, due to low sample size per treatment, I was not able to perform statistical analyses on either crab behavior or predation. Trial length was initially 24 h, but due to 100% tclam predation in the first few trials, trial length was shortened to 12 hours in an attempt to reduce clam predation. However, crabs in the 12-h trials did not eat at all. Thus, crab behavior and clam predation varied significantly depending on trial length, and crab sample size of each treatment was too small to analyze data from long trials and short trials separately. Thus, long and short trials were combined to examine trends in the data. In addition, both pH levels were combined to examine trends by salinity alone, and both salinity levels were combined to examine trends by pH alone. I visualized trends in data on crab behaviors between the two pH and two salinity levels using stacked bar graphs for each parameter.
3. RESULTS
3.1 Seawater Chemistry
In 2018, salinity, temperature, pH, and DO remained relatively stable over the 11 weeks
(Figures 1A-D), and the nominal treatments were maintained. DIC, from weeks 1, 2, and 11, increased over time in both acidified treatments, but was highest in the low-pH, high-salinity treatment (Figure 2A). DIC did not change over time in the non-acidified treatments. Carbonate alkalinity (CA) was essentially the inverse trend of DIC, with the acidified treatments varying little over time (Figure 2B). CA increased over time in the non-acidified treatments, and was greatest overall in the non-acidified, high-salinity treatment.
14
In 2019, salinity, temperature, pH, and DO remained relatively stable over the 10 weeks
(Figure 3). Ambient pH was overall lower in crab tanks than clam tanks by ~0.5 pH units but acidified pH remained on average between 7.0 and 7.3 in both crab and clam tanks (Figure 3A-
B). Low- and high-salinity levels were maintained over the duration of the experiment in both crab and clam tanks (Figure 3C-D). DO was much lower in the crab tanks than in the clam tanks, mostly due to the crab tanks being much larger and thus harder to aerate (Figure 3E-F).
In the larger tanks, water circulation was more difficult to control, however, DO never fell below
5 mg/L. Temperature was higher in the crab tanks than in the clam tanks during the first half of the experiment, but decreased to the same temperature as the clam tanks in the second half
(Figure 3G-H). Regarding DIC and CA, only weeks 8 and 10 were analyzed, thus trends spanning the duration of the experiment could not be determined (Figure 4A-D). DIC decreased slightly in the clam tanks between week 8 and 10, but increased in the crab tanks. CA remained stable in the clam tanks between weeks 8 and 10, but increased in the crab control treatment.
3.2 Single-species impacts
3.2.1 Crab carapace strength
In 2018, blue crab carapace strength was best explained by the additive model (g12) of crab size, exposure time (ET), and pH, which had the best weighted probability of 0.979 and the best fit (Table 7). All other models had weighted probabilities < 0.02. The parameter estimates from model g12 (Table 8) were used to generate carapace strength as:
푦 = 47.56 + 0.51푥1 − 0.28푥2 + 15.76푥4 (2) where x1 = crab size, x2 = exposure time, and x4 = pH. This equation was used to graph carapace strength by size (Figure 5) and exposure time (Figure 6), with data color-coordinated by high and low pH. Carapace strength increased with crab size, decreased with ET (the carapace deteriorated as it aged), and was overall stronger in low pH compared to high pH.
With increased sample size in 2019, blue crab carapace strength was best explained by the interactive model (g7) with salinity and pH, which had the best weighted probability of 0.614
15 and the best fit (Table 9). The second- and third-best models included size plus the interaction of salinity and pH and exposure time plus the interaction of salinity and pH (models g9 and g10, respectively). These models explained approximately the same percent deviance; however, the weighted probabilities were significantly weaker than for model g7. The parameter estimates from model g7 (Table 10) were used to generate carapace strength as:
푦 = 85.208 − 1.637푥3 − 21.542푥4 + 26.692푥6 (3) where x3 = salinity, x4 = pH and x6 = salinity*pH. Mean carapace strength was then visualized by treatment (Figure 7). Carapace strength was weakest in the low-pH, high-salinity treatment and strongest in the low-pH, low-salinity treatment. The high-pH treatments had similar mean carapace strength regardless of salinity.
3.2.2 Crab claw pinch force
In 2018, blue crab claw pinch force was best explained by the additive model (g7) of crab size and pH, which had the second-best weighted probability and model fit (Table 11). The best model, g2, only included crab size. I wanted, however, to examine the relative importance of other variables and therefore chose to evaluate model g7. Model g7 explained significantly more deviance than model g2. Thus, parameter estimates from model g7 (Table 12) were used to generate claw pinch force as:
푦 = −7.110 + 0.234푥1 − 4.213푥4 (4) where x1 = crab size and x4 = pH. This equation was used to graph claw pinch force by carapace width (Figure 8) with data color-coordinated by pH levels. Pinch force increased with size but was weaker in low pH.
In 2019, blue crab claw pinch force was best explained by the additive model (g6) of crab size, pH, and sex, which had the best weighted probability and the best model fit (Table 13).
The second- and third-best models by weight and model fit included model g2 with size alone, and model g3, with pH plus sex. Neither model, however, had a weighted probability or
16 explained nearly as much deviance as model g6. Thus, parameter estimates from model g6
(Table 14) were used to generate claw pinch force as:
푦 = −11.05 + 0.291푥1 − 3.650푥4 + 5.147푥5 (5) where x1 = crab size, x4 = pH and x5 = sex. This equation was used to graph claw pinch force by carapace width (Figure 9) and sex (Figure 10). Pinch force increased with crab carapace width and was overall weaker in low-pH treatments, as in 2018. Moreover, male crabs had a stronger pinch than female crabs.
3.2.3 Clam growth
Clam growth in 2018 was best explained by pH alone (model g3) with a model weight of
0.254 (Table 15). The second-best model was the additive model g7 with pH and salinity, which explained slightly more deviance but had a lower model weight and an additional parameter.
Analysis of deviance revealed that the model with more parameters, model g7, did not significantly reduce the deviance explained. Parameter estimates from model g3 (Table 16) were used to generate clam growth as:
푦 = −0.000 − 0.0153푥2 (6) where x2 = pH. This equation was used to graph the mean slopes of clam growth by treatment
(Figure 11A). Clams in low-pH treatments had negative growth, while clams in high-pH treatments had positive growth, regardless of salinity.
In 2019, clam growth was best explained by the additive model g8 with initial size, pH and salinity (wi = 0.254) (Table 17). This model was closely followed by model g13, with pH plus the interaction of size and salinity (wi = 0.236). Despite that model g13 explained slightly more deviance, analysis of deviance showed that the additional parameters did not significantly reduce deviance. Parameter estimates from model g8 (Table 18) were used to generate clam growth as:
푦 = 0.3213 − 0.0247푥1 − 0.0448푥2 − 0.0490푥3 (8)
17 where x1 = initial shell length, x2 = pH, and x3 = salinity. This equation was used to calculate the mean slopes of clam growth over time by treatment (Figure 11B). Overall, the most significant difference in clam growth was between the low-pH, low-salinity treatment and the control treatment, as expected. A positive slope was only seen in the high-pH, high-salinity treatment.
Salinity played a greater role in 2019 than in 2018.
3.2.4 Clam survival
In 2018, clam survival, similar to clam growth, was best explained by pH alone (model g3) with a model weight of 0.298 (Table 19). The second-best model was additive with initial length and pH, which explained slightly more deviance but had a lower model weight.
Parameter estimates from model g3 (Table 20) were used to generate clam survival as:
푦 = 0.857 − 0.196푥2 (7) where x2 = pH. This equation was used to graph clam survival by treatment (Figure 12A). As expected, clams in the low-pH, low-salinity treatment had the lowest survival percentage, while those in the control treatment had the highest survival.
Clam survival in 2019 was more influenced by salinity than by pH. The best model was additive with initial length and salinity (model g6, Table 21). The second-best model was with salinity alone (model g4). Only the third-best model (g8) included pH, along with salinity and starting length. Parameter estimates from model g6 (Table 22) were used to generate clam survival as:
푦 = −0.409 + 0.111푥1 − 0.3556푥3 (9) where x1 = initial shell length and x3 = salinity. This equation was used to graph clam survival by treatment (Figure 12B). As in 2018, survival was lowest in the low-pH, low-salinity treatment and highest in the control treatment, although the difference between the survival rates was greater in 2019 than in 2018.
3.2.5 Clam shell structure and ridge rugosity (2018)
18
The most stressful treatment (low pH and low salinity) had the highest mean damage index (DI) score (± 0.2SE) (as defined above in section of 2.2.3), while the control treatment had the lowest mean DI of 0.5 (± 0.2 SE). The second-most-stressful treatment was with low pH and high salinity (DI = 1.5 ± 0.2 SE) and the third-most-stressful treatment was with high pH and low salinity (DI = 1.2 (± 0.3 SE). As expected, shells in the acidified treatments tended to have more damage than shells in the non-acidified treatments, and shells in the low-salinity treatments tended to appear more damaged than shells in the high-salinity treatments (Figure 13).
Rugosity (shell deterioration) was best explained by pH alone (model g3), followed by the additive model of pH and salinity (model g3) (Table 23). Parameter estimates from model g3
(Table 24) were used to generate clam ridge rugosity as:
푦 = 1.200 + 0.692푥2 (10) where x2 = pH. Rugosity was greater in the acidified treatments and highest in the most- stressful treatment (Figure 14). When comparing salinity treatments, rugosity was greater in low salinity.
3.3 Filmed predator-prey trials (2019)
Salinity tended to have a minor impact on crab behavior, as visualized in trends revealed in stacked bar graphs. Crabs tended to spend more time eating and burying in high-salinity treatments, but more time moving in low-salinity treatments (Figure 15 A). The number of trials in high salinity, however, was more than double the number of trials in low salinity (because of differential survival and availability of clams after treatment exposure). This apparent trend could thus be the result of higher sample size in high salinity. pH did not have an impact on crab behavior, and time spent eating was not different in high- and low-pH trials (Figure 15 B).
Clam mortality during the trials was directly linked to trial length; clams in longer trials
(24 h) were consumed much more than clams in shorter trials (12 h). Trials of different lengths were not evenly distributed among treatments. Furthermore, the trends toward increased crab eating behavior with high salinity levels were not reflected in clam mortality. When averaged by
19 treatment, crabs consumed 3.3 (± 1.5 SE) clams in the low-pH, high-salinity treatment (6 trials),
3.5 (± 3.5 SE) clams in the low-pH, low-salinity treatment (2 trials), 7 (± 0 SE) clams in the high- pH, low-salinity treatment (2 trials), and 2.5 (± 1.31 SE) clams in the control high-pH, high- salinity treatment (4 trials).
4. DISCUSSION
In a multi-year study, with multiple stressors, both juvenile crabs and clams were most impacted by pH as compared to salinity or a combination of multiple stressors. In both years, crab claw pinch force was weaker when crabs were exposed to acidified conditions; however, carapace strength yielded mixed results. Blue crabs are strong osmoregulators and strong acid- base regulators, whereas hard clams are osmoconformers but weak acid-base regulators. Clam growth, survival, exterior shell structure and ridge rugosity were negatively impacted by both pH and low salinity. Yet despite the impacts on species separately, predator-prey trials indicate no alteration of the relationship due to salinity or low pH.
4.1 Crab carapace strength and pinch force
After 10 – 11 weeks in the experimental conditions of low and high pH and low and high salinity in 2018 and 2019, adult blue crab carapace strength varied between the years. In 2018, crabs had a stronger carapace in low-pH treatments, regardless of salinity, whereas in 2019, crabs had a stronger carapace in high-pH and low-salinity treatments. Low sample size could have contributed to the 2018 result in low pH, but many studies indicate similar mixed results for crabs at the juvenile and adult life stages. For example, there was net positive calcification in adult blue crabs (n = 8) exposed to pH 7.31 for 60 days (Ries et al. 2009), corroborating the observations in 2018. In contrast, when juvenile blue crabs (n = 160) were exposed to pH 6.7, also for 60 days, there was no significant impact of acidification (Glandon & Miller 2017). Given that pH is calculated as the log of hydrogen ion (H+) concentration, these small differences in pH represent large differences in H+ concentrations. In shorter-term studies, other brachyuran
20 crabs acclimated to acidification (Spicer et al. 2007, Pane & Barry 2007, Small et al. 2010) or the carapace was not impacted by acidification (Amaral et al. 2012, Coffey et al. 2017).
Regarding salinity, blue crabs are strong osmoregulators; therefore, salinity was not expected to be detrimental to crabs in this study. Yet carapace strength was greater in low salinity than high salinity in 2019, a result not measured in previous salinity experiments on blue crabs (e.g., Whitely et al. 2001, Ramaglia et al. 2018). This could be due to the salinity levels from which the crabs were collected. Individuals were collected from the York River, in which salinity levels range from 12 – 20 ppt (CBNERR 2018). Crabs, therefore, would have been more frequently exposed to the to the experimental low salinity (16 ppt) versus the high salinity (30 ppt) prior to capture. Although I could not find previous literature on the carapace strength of any decapod in varying salinities, many physiological studies exist. In one short-term (24 h) physiological study, blue crabs compensated for low salinity (8 ppt, Curtis & McGaw 2010), but in other short-term physiological studies with brachyuran crabs, low salinity negatively altered acid-base regulation in the Chinese mitten crab Eriocheir sinensis (10 ppt, Whiteley et al. 2001) and the swimming crab Callinectes danae (20 ppt, Ramaglia et al. 2018). Multiple stressors in addition to salinity can be detrimental to crabs, as in the latter experiment, C. danae lost osmoregulatory abilities at low salinities when also exposed to pH 7.3 for 3 days. Long-term studies with pH and salinity on crustaceans are scarce.
Finally, regarding exposure time (ET), blue crab carapace strength decreased over time in 2018, but ET did not affect carapace strength in 2019. A decrease in carapace strength over time biologically makes sense since the carapace is not maintained once it hardens post-molt.
In a previous study, juvenile lobsters exposed to acidification exhibited altered intermolt period lengths (McLean et al. 2018). In my research, most crabs did not molt until the end of the 2018 experiment, heavily skewing the data towards long exposure times. As ET was not as significant in 2019 and model deviances were low, there could have been other factors I did not take into account. For example, experimental temperature was greater in 2019 than 2018 (28 vs. 24,
21 respectively), potentially influencing blue crab molting frequency (Glandon & Miller 2017).
Carapace strength is a vital defense for juvenile crabs from predators, including adult blue crabs
(Dittel et al. 1995, Hines & Ruiz 1995). Future blue crab populations may be able to tolerate low pH, because they do not change their intermolt period or growth (Glandon et al. 2017), but the energetic trade-offs of maintaining the carapace are under-studied.
Claw pinch force was significantly reduced by acidification in both years, regardless of salinity. This trend is corroborated by studies where low pH negatively impacted the muscle length and strength of green crabs (Landes & Zimmer 2012), the endocuticle microhardness in juvenile red and blue king crabs (Coffey et al. 2017), and an increase in [Mg2+] in velvet swimming crab chelae, an indication of reduced chelae strength (Small et al. 2010).
Ecologically, decreased claw strength could be detrimental to the future of blue crab populations. Individuals could become unable to properly consume their typical prey, such as clams (Hines et al. 1990), and be unable to defend themselves from predators. Few studies have examined the ecological impacts of OA on juvenile blue crabs (but see Lane et al. 2013,
Glandon & Miller 2017, Glandon et al. 2018), which is where my work is useful in filling knowledge gaps.
4.2 Clam growth, survival, shell structure, and ridge rugosity
Concerning impacts of pH and salinity on juvenile hard clams, pH alone had the greatest negative impact on both growth and survival in 2018. Low salinity also negatively impacted clam survival, but this effect was not as strong as that for pH. In 2019, the additive effects of pH and salinity negatively impacted clam growth and survival more than pH alone. Salinity contributed more significantly than in 2018, with survival being high in both high-salinity treatments, relative to low-salinity treatments. Because clams are osmoconformers (Anderson & Prosser 1953,
DuPaul & Webb 1974) and weak acid-base regulators (Melzner et al. 2009), these results are not surprising. Although low salinity had a more noticeable impact, low pH also negatively impacted juvenile hard clam survival in a previous 21-week study involving pH levels 8.2, 8.1,
22 and 7.7 and salinity levels 16 and 32 (Dickinson et al. 2013). To determine the effects of even lower pH, I replicated the salinity levels from Dickinson et al. (2013), but at a lower pH. This allowed me to determine lethal and sublethal effects of lower pH. Early-life-stage hard clams also experienced reduced growth in combined low-pH and low-dissolved-oxygen treatments
(Gobler et al. 2014), thus multiple stressors can exacerbate effects of OA. Finally, initial shell length of clams played a role in survival, with smaller clams having a lower survival rate. Shells of younger clams are not as developed as those of adult clams; thus, young clams have difficulty overcoming net dissolution (Waldbusser et al. 2010). OA has also proved detrimental in other marine calcifiers. Mediterranean mussels exposed to low pH experienced shell dissolution
(Bressan et al. 2014), juvenile eastern oysters exposed to low pH and low salinity had significant mortality (Dickinson et al. 2012), and shells of pearl oysters were weakened under acidification (Welladsen et al. 2010).
As seen in SEM images, including growth ridges, shells in acidified treatments were both qualitatively and quantitatively more damaged than shells in high-pH treatments. In addition, shells in low-salinity treatments were more damaged than shells in high-salinity treatments and the most damaged shells were in the combined low-pH and low-salinity treatment. Shell dissolution under such stressful conditions could make clams more vulnerable to predation
(Vermeij 1987). If future salinity levels in the Chesapeake Bay decrease because of increased storm activity, as predicted by some (Jay et al. 2018), the stressful conditions in this experiment could become a reality, and clams could be compromised.
4.3 Predator-prey interactions
Filmed predator-prey trials of crab behavior and predation on clams indicated that pH had little impact on crab behavior after the 10 weeks of exposure. Given that other brachyuran crabs acclimated to low pH in previous studies, it is not surprising that OA does not impact blue crab behavior. However, if blue crab pinch force and carapace strength are negatively impacted, the question becomes, for how long can an individual maintain homeostasis and normal
23 behavior? Behavior may eventually change as the energetic trade-offs become too great, such as muscle wastage or carapace dissolution because of the high costs of increased calcification
(Spicer et al. 2007, Wood et al. 2008). This question remains understudied, particularly in blue crabs, and is an area for further research. OA increased the foraging time and decreased consumption rates of the brown crab Cancer pagurus on blue mussels Mytilus edulis (Wang et al. 2018), as well as for the crab Charybdis japonica on a number of prey (Wu et al. 2017).
Alternatively, if both predator and prey are altered (as in the present study), the result could be no net impact on their relationship (Landes & Zimmer 2012, Glaspie et al. 2017). In this experiment, clam shells were degraded and crab claw pinch force was weaker, suggesting that under future acidified conditions, there might be no net alteration in the crab-clam predator-prey relationship.
Behavior analyses also indicate that crabs tend to spend more time eating in high- salinity treatments. If storm activity increases and future salinity levels decrease in the
Chesapeake Bay, as some predict (Jay et al. 2018), and behavioral trends bear out over the long term, crabs could consume less, decreasing their nutrient and energy intake and potentially leading to lower growth or even starvation. This is corroborated by a previous study where blue crabs ate significantly less in low salinity than high salinity (Stickle et al. 2007). Because blue crabs expend more energy to maintain osmotic balance in low salinities (Guerin & Stickle 1992), they may not have enough energy for foraging and growth, which could affect future blue crab populations.
5. SUMMARY
As the Chesapeake Bay undergoes climatic changes, stressors including OA and greater fluctuations in salinity could pose problems for marine calcifying species. Although blue crab may be able to withstand future stressful conditions, the energetic trade-offs may be required (e.g., maintaining carapace strength versus muscle strength). For sedentary species
24 like hard clams and other bivalves, however, the detrimental effects of OA and increased precipitation could threaten future populations. Understanding how OA interacts with other stressors to affect species responses is necessary for future management of exploited species, particularly under climate change.
25
LITERATURE CITED
Aldridge JB, Cameron JN (1982) Gill morphometry in the blue crab, Callinectes sapidus Rathbun (Decapoda Brachyura). Crustaceana 43:297-305
Amaral V, Cabral HN, Bishop MJ (2012) Effects of estuarine acidification on predator-prey interactions. Aquat Ecol 46:119–127 doi: 10.3354/meps09487
Anderson JD, Prosser CL (1953) Osmoregulation capacity in populations occurring in different salinities (abstr.). Biol Bull 105:369
Anderson D (2008) Model based inference in the life sciences: a primer on evidence. New York, NY: Springer Science and Business Media
Baumann H, Smith EM (2017) Quantifying Metabolically Driven pH and Oxygen Fluctuations in US Nearshore Habitats at Diel to Interannual Time Scales. Estuaries Coast 41:1102– 1117 https://doi.org/10.1007/s12237-017-0321-3
Bibby R, Cleall-Harding P, Rundle S, Widdicombe S, Spicer J (2007) Ocean acidification disrupts induced defences in the intertidal gastropod Littorina littorea. Biol Lett 3:699– 701 https://doi.org/10.1098/rsbl.2007.0457
Blundon JA, Kennedy VS (1982) Mechanical and behavioral aspects of blue crab, Callinectes sapidus (Rathbun), predation on Chesapeake Bay bivalves. J Exp Mar Biol Ecol 65:47- 65 https://doi.org/10.1016/0022-0981(82)90175-7
Bressan M, Chinellato A, Munari M, Matozzo V, Manci A, Marčeta T, Finos L, Moro I, Pastore P, Badocco D, Marin MG (2014) Does seawater acidification affect survival, growth and shell integrity in bivalve juveniles? Mar Environ Res 99:136–148 https://doi.org/10.1016/j.marenvres.2014.04.009
Bruno JF, O’Connor MI (2005) Cascading effects of predator diversity and omnivory in a marine food web. Ecol Lett 8:1048–1056 https://doi.org/10.1111/j.1461-0248.2005.00808.x
Burnett LE, McMahon BR (1985) Facilitation of CO2 excretion by carbonic anhydrase located on the surface of the basal membrane of crab gill epithelium. Respir Physiol 62(3):341-8 doi: 10.1016/0034-5687(85)90089-1
Burnett LE (1984) CO2 excretion across isolated perfused crab gills: facilitation by carbonic anhydrase. Amer Zool 24:253–264
Burnham KP, Anderson DR (2002) Model selection and multimodel inference: a practical information-theoretic approach. 2nd ed. New York, NY: Springer Science and Business Media
Cai W, Wang Y (1998) The chemistry, fluxes, and sources of carbon dioxide in the estuarine waters of the Satilla and Altamaha Rivers, Georgia. Limnol Oceanogr 43:657–668 https://doi.org/10.4319/lo.1998.43.4.0657
Cai W, Huang W-J, Luther GW, Pierrot D, Li M, Testa J, Xue M, Joesoef A, Mann R, Brodeur J, Xu Y-Y, Chen B, Hussain N, Waldbusser GG, Cornwell J, Kemp WM (2017) Redox
26
reactions and weak buffering capacity lead to acidification in the Chesapeake Bay. Nat Commun 8:369 https://doi.org/10.1038/s41467-017-00417-7
Caldeira K, Wickett ME (2003) Oceanography: anthropogenic carbon and ocean pH. Nature 425:365 https://doi.org/10.1038/425365a
Carter LM, Jones JW, Berry L, Burkett V, Murley JF, Obeysekera J, Schramm PJ, Wear D (2014) Chapter 17: Southeast and the Caribbean: Climate Change Impacts in the United States: The Third National Climate Assessment, in: Melillo, J.M., Richmond, T.C., Yohe, G.W. (Eds.), U.S. Global Change Research Program. U.S. Global Change Research Program, pp. 396–417. https://doi.org/10.7930/J0N-P22CB
Chamberlin TC (1890) The method of multiple working hypotheses. Science 15: 92–96 (reprinted in Science 148:754–759 [1965])
Coffey WD, Nardone JA, Yarram A, Long WC, Swiney KM, Foy RJ, Dickinson GH (2017) Ocean acidification leads to altered micromechanical properties of the mineralized cuticle in juvenile red and blue king crabs. J Exp Mar Biol Ecol 495:1-12 https://doi.org/10.1016/j.jembe.2017.05.011
Colen C Van, Debusschere E, Braeckman U, Gansbeke D Van, Vincx M (2012) The Early Life History of the Clam Macoma balthica in a High CO2 World. PLoS One 7:1–8 https://doi.org/10.1371/journal.pone.0044655
Crenshaw MA (1980) Mechanisms of shell formation and dissolution. In: D.C. Rhoads and R.A. Lutz (Eds.), Skeletal Growth of Aquatic Organisms: Biological Records of Environmental Change. Plenum Press, New York, pp. 115-132
Curtis DL, McGaw IJ (2010) Respiratory and digestive responses of postprandial Dungeness crabs, Cancer magister, and blue crabs, Callinectes sapidus, during hyposaline exposure. J Comp Physiol B 180:89–198 https://doi.org/10.1007/s00360-009-0403-z
Denman K, Christian JR, Steiner N, Pörtner H-O, Nojiri Y (2011) Potential impacts of future ocean acidification on marine ecosystems and fisheries: current knowledge and recommendations for future research. ICES J Mar Sci 68:1019–1029 https://doi.org/10.1093/icesjms/fsr074
Dickinson GH, Ivanina AV, Matoo OB, Pörtner H-O, Lannig G, Bock C, Beniash E, Sokolova IM (2012) Interactive effects of salinity and elevated CO2 levels on juvenile eastern oysters, Crassostrea virginica. J Exp Biol 215:29–43 https://doi.org/10.1242/jeb.061481
Dickinson GH, Matoo OB, Tourek RT, Sokolova IM, Beniash E (2013) Environmental salinity modulates the effects of elevated CO2 levels on juvenile hard-shell clams, Mercenaria mercenaria. J Exp Biol 216:2607–2618 https://doi.org/10.1242/jeb.082909
Dittel A, Hines A, Ruiz G, Gregory M, Ruffin K (1995) Effects of shallow water refuge on behavior and density-dependent mortality of juvenile blue crabs in Chesapeake Bay. Bull Mar Sci 57:902–916
27
Dodd LF, Grabowski JH, Piehler MF, Westfield I, Ries JB (2015) Ocean acidification impairs crab foraging behaviour. Proc Biol Sci 282:20150333 https://doi.org/10.1098/rspb.2015.0333
Dodd L (2012) Ocean acidification affects oyster-mud crab predator-prey dynamics. 41st Benthic Ecology Meeting, Norfolk, VA, March 21-24, 2012. Meeting abstracts
Doney SC, Schimel DS (2007) Carbon and Climate System Coupling on Timescales from the Precambrian to the Anthropocene. Annu Rev Environ Resour 32:31–66 https://doi.org/10.1146/annurev.energy.32.041706.124700
Doney SC, Fabry VJ, Feely RA, Kleypas JA (2009) Ocean acidification: the other CO2 problem. Ann Rev Mar Sci 1:169–192
Donohue P, Calosi P, Bates AH, Laverock B, Rastrick S, Mark FC, Strobel A, Widdicombe S (2012) Impact of exposure to elevated pCO2 on the physiology and behaviour of an important ecosystem engineer, the burrowing shrimp Upogebia deltaura. Aquat Biol 15:73-86. https://doi.org/10.3354/ab00408
Duarte CM (2000) Marine biodiversity and ecosystem services: an elusive link. J Exp Mar Biol Ecol 250:117–131 https://doi.org/10.1115/1.4026224
DuPaul WD, Webb KL (1974) Salinity-induced changes in the alanine and aspartic aminotransferase activity in three marine bivalve molluscs. Arch Int Physiol Biochim 82:817-822
Eggleston DB, Lipcius RN, Hines AH (1992) Density-dependent predation by blue crabs upon infaunal clam species with contrasting distribution and abundance patterns. Mar Ecol Prog Ser. 85:55–68 https://doi.org/10.3354/meps085055
Ellis RP, Bersey J, Rundle SD, Hall-Spencer JM, Spicer JI (2009) Subtle but significant effects of CO2 acidified seawater on embryos of the intertidal snail, Littorina obtusata. Aquat Biol 5:41-48. https://doi.org/10.3354/ab00118
Fabricius KE, Langdon C, Uthicke S, Humphrey C, Noonan S, De’ath G, Okazaki R, Muehllehner N, Glas MS, Lough JM (2011) Losers and winners in coral reefs acclimatized to elevated carbon dioxide concentrations. Nat Clim Chang 1:165–169 https://doi.org/10.1038/nclimate1122
Findley AM, Belisle BW, Stickle WB (1978) Effects of salinity fluctuactions on the respiration rate of the southern oyster drill Thais haemastoma and the blue crab Callinectes sapidus. Mar Biol 49:59-67
Foyle TP, Hurley GV, Taylor DM (1989) Field testing shell hardness gauges for the snow crab fishery. Canadian Industry Report of Fisheries and Aquatic Sciences 193:1–38.
Gazeau F, Quiblier C, Jansen JM, Gattuso J-P, Middelburg JJ, Heip CHR (2007) Impact of elevated CO2 on shellfish calcification. Geophys. Res Lett 34:L07603 https://doi.org/10.1029/2006GL028554
28
Glandon HL, Miller TJ (2017) No effect of high pCO2on juvenile blue crab, Callinectes sapidus, growth and consumption despite positive responses to concurrent warming. ICES J Mar Sci 74:1201–1209 https://doi.org/10.1093/icesjms/fsw171
Glandon HL, Kilbourne KH, Schijf J, Miller TJ (2018) Counteractive effects of increased temperature and pCO2 on the thickness and chemistry of the carapace of juvenile blue crab, Callinectes sapidus, from the Patuxent River, Chesapeake Bay. J Exp Mar Biol Ecol 498:39–45 https://doi.org/10.1016/j.jembe.2017.11.005
Glaspie CN, Longmire K, Seitz RD (2017) Acidification alters predator-prey interactions of blue crab Callinectes sapidus and soft-shell clam Mya arenaria. J Exp Mar Biol Ecol 489:58– 65 https://doi.org/10.1016/j.jembe.2016.11.010
Giltz SM, Taylor CM (2017) Reduced growth and survival in the larval blue crab Callinectes sapidus under predicted ocean acidification. J Shellfish Res 36(2): 481-485.
Gobler CJ, DePasquale EL, Griffith AW, Baumann H (2014) Hypoxia and Acidification Have Additive and Synergistic Negative Effects on the Growth, Survival, and Metamorphosis of Early Life Stage Bivalves. PLoS ONE 9(1):e83648 https://doi.org/10.1371/journal.pone.0083648
Guerin JL, Stickle WB (1992) Effects of salinity gradients on the tolerance and bioenergetics of juvenile blue crabs (Callinectes sapidus) from waters of different environmental salinities. Mar Biol 114:391–396 https://doi.org/10.1007/BF00350029
Hamilton PV, Nishimoto RT, Halusky JG (1976) Cheliped laterality in Callinectes sapidus (Crustacea: Portunidae) Biol Bull 150:393-401
Henry RP (1988) Multiple functions of carbonic anhydrase in the crustacean gill. J Exp Zool 248:19-24 doi:10.1002/jez.1402480104
Hilton TW, Najjar RG, Zhong L, Li M (2008) Is there a signal of sea-level rise in Chesapeake Bay salinity? J Geophys Res 113:1–12 https://doi.org/10.1029/2007JC004247
Hines A, Ruiz G (1995) Temporal variation in juvenile blue crab mortality: nearshore shallows and cannibalism in Chesapeake Bay. Bull Mar Sci 57:884–901
Hines AH, Haddon PJ, Miklas JJ, Wiechert LA, Haddon AM (1987) Estuarine invertebrates and fish: Sampling design and constraints for long-term measurements of population dynamics, New Approaches to Monitoring Aquatic Ecosystems
Hines AH, Haddon AM, Wiechert LA (1990) Guild structure and foraging impact of blue crabs and epibenthic fish in a subestuary of Chesapeake Bay. Mar Ecol Prog Ser 67:105–126 https://doi.org/10.3354/meps067105
Jay A, Reidmiller DR, Avery CW, Barrie D, DeAngelo BJ, Dave A, Dzaugis M, Kolian M, Lewis KLM, Reeves K, Winner D (2018) Overview. In Impacts, Risks, and Adaptation in the United States: Fourth National Climate Assessment, Volume II [Reidmiller, D.R., C.W. Avery, D.R. Easterling, K.E. Kunkel, K.L.M. Lewis, T.K. Maycock, and B.C. Stewart (eds.)]. U.S. Global Change Research Program, Washington, DC, USA, pp. 33–71. doi: 10.7930/NCA4.2018.CH1
29
Johnson PT (1980) Histology of the blue crab, Callinectes sapidus: A model for the Decapoda. Praeger Scientific, New York 440 p.
King EN (1965) The oxygen consumption of intact crabs and excised gills as a function of decreased salinity. Comp Biochem Physiol 15:93-102
Kroeker KJ, Kordas RL, Crim R, Hendriks IE, Ramajo L, Singh GS, Duarte CM, Gattuso JP (2013) Impacts of ocean acidification on marine organisms: Quantifying sensitivities and interaction with warming. Glob Chang Biol 19:1884–1896 https://doi.org/10.1111/gcb.12179
Landes A, Zimmer M (2012) Acidification and warming affect both a calcifying predator and prey, but not their interaction. Mar Ecol Prog Ser 450:1–10 https://doi.org/10.3354/meps09666
Lane HA, Paynter KT, Miller TJ (2013) Acidification impacts growth per molt and hardening time of juvenile blue crab Callinectes sapidus. 2013 Coastal and Estuarine Research Federation Conference Abstract Book, p. 129 http://www.sgmeet.com/cerf2013/files/CERF2013-Abstract-Book.pdf
Lipcius RN, Latour RJ (2006) Food Web Interactions and Modeling. Fish Ecosyst Plan Chesap Bay 103–144
Lipcius RN, Van Engel WA (1990) Blue Crab Population Dynamics in Chesapeake Bay: Variation in Abundance (York River, 1972-1988) and Stock-Recruit Functions. Bull Mar Sci 46:180–194
Lipcius RN, Eggleston DB, Heck KLJ, Seitz RD, Montfrans J Van (2007) Post-settlement abundance, survival, and growth of postlarvae and young juvenile blue crabs in nursery habitats, in: Kennedy, V., Cronin, L. (Eds.), The Blue Crab Callinectes Sapidus. Maryland Sea Grant College, College Park (MD), pp. 535–564
Mann R, Harding JM, Southworth MJ, Wesson JA (2005) Northern quahog (hard clam) Mercenaria mercenaria abundance and habitat use in Chesapeake Bay. J Shellfish Res 24:509–516 https://doi.org/10.2983/0730-8000(2005)24[509:NQHCMM]2.0.CO;2
Mansour RA, Lipcius RN (1991) Density-dependent foraging and mutual interference in blue crabs preying upon infaunal clams. Mar Ecol Prog Ser 72:239–246 https://doi.org/10.3354/meps072239
MATLAB (2020) version 9.8.0.1417392 (R2020a). Natick, Massachusetts: The MathWorks Inc.
McLean EL, Katenka NV, Seibel BA (2018) Decreased growth and increased shell disease in early benthic phase Homarus americanus in response to elevated CO2. Mar Ecol Prog Ser 596:113-126 https://doi.org/10.3354/meps12586
MdOA. 2015. Task Force to Study the Impact of Ocean Acidification on State Waters: Report to the Governor and the Maryland General Assembly. 46 pp.
Melzner F, Gutowska MA, Langenbuch M, Dupont S, Lucassen M, Thorndyke MC, Bleich M, Pörtner H-O (2009) Physiological basis for high CO2 tolerance in marine ectothermic
30
animals: pre-adaptation through lifestyle and ontogeny? Biogeosciences Discuss 6:4693–4738 https://doi.org/10.5194/bgd-6-4693-2009
Menge BA, Sutherland JP (1987) Community Regulation: Variation in Disturbance, Competition, and Predation in Relation to Environmental Stress and Recruitment. Am Nat 130 https://doi.org/10.1086/284741
Menge BA (1995) Indirect effects in marine rocky intertidal interaction webs: Patterns and importance. Ecol Monogr 65:21-74
Metzger R, Sartoris FJ, Langenbuch M, Pörtner H-O (2007) Influence of elevated CO2 concentrations on thermal tolerance of the edible crab Cancer pagurus. J Therm Biol 32:144–151 https://doi.org/10.1016/j.jtherbio.2007.01.010
Micheli F (1997) Effects of Predator Foraging Behavior on Patterns of Prey Mortality in Marine Soft Bottoms. Ecol Monogr 67:203–224
Miller CA, Waldbusser GG (2016) A post-larval stage-based model of hard clam Mercenaria mercenaria development in response to multiple stressors: Temperature and acidification severity. Mar Ecol Prog Ser 558:35–49 https://doi.org/10.3354/meps11882
Miller TJ, Martell SJD, Bunnell DB, Davis G, Fegley L, Sharov A, Bonzek C, Hewitt D, Hoenig J, Lipcius RN (2005) Stock assessment of blue crab in Chesapeake Bay. Technical Report Series No. TS-487-05 of the University of Maryland Center for Environmental Science, College Park, MD.
Miller AW, Reynolds AC, Sobrino C, Riedel GF (2009) Shellfish face uncertain future in high CO2 world: Influence of acidification on oyster larvae calcification and growth in estuaries. PLoS One 4. https://doi.org/10.1371/journal.pone.0005661
Najjar RG, Pyke CR, Adams MB, Breitburg D, Hershner C, Kemp M, Howarth R, Mulholland MR, Paolisso M, Secor D, Sellner K, Wardrop D, Wood R (2010) Potential climate- change impacts on the Chesapeake Bay. Estuar Coast Shelf Sci 86:1–20 https://doi.org/10.1016/j.ecss.2009.09.026
Narita D, Rehdanz K, Tol RSJ (2012) Economic costs of ocean acidification: A look into the impacts on global shellfish production. Clim Change 113:1049–1063 https://doi.org/10.1007/s10584-011-0383-3
National Oceanic and Atmospheric Administration (NOAA). 2019. Landings. https://foss.nmfs.noaa.gov/apexfoss/f?p=215:200:1228787575241::NO:::.
Neufeld GJ, Holliday CW, Pritchard JB (1980) Salinity adaption of gill Na, K‐ATPase in the blue crab, Callinectes sapidus. J Exp Zool Part A 211:215-224
Pane EF, Barry JP (2007) Extracellular acid-base regulation during short-term hypercapnia is effective in a shallow-water crab, but ineffective in a deep-sea crab. Mar Ecol Prog Ser 334:1–9 https://doi.org/10.3354/meps334001
31
Parker LM, Ross PM, O’Connor WA, Pörtner H-O, Scanes E, Wright JM (2013) Predicting the response of molluscs to the impact of ocean acidification. Biology 2:651–92 https://doi.org/10.3390/biology2020651
Pearl HW (2006) Assessing and managing nutrient-enhanced eutrophication in estuarine and coastal waters: Interactive effects of human and climatic perturbations. Ecol Eng 26:40- 54
Peterson CH (1983) A concept of quantitative reproductive senility: application to the hard clam, Mercenaria mercenaria (L.). Oecologia 58:164-168
Pörtner H-O, Langenbuch M, Reipschläger A (2004) Biological impact of elevated ocean CO2 concentrations: lessons from animal physiology and earth history. J Oceanogr 60:705- 718
R Core Team (2017) R: A Language and Environment for Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Retrieved from https://www.R- project.org/
RStudio Team (2017) RStudio: Integrated Development Environment for R. Boston, MA: RStudio, Inc. Retrieved from http://www.rstudio.com/
Ramaglia AC, de Castro LM, Augusto A (2018) Effects of ocean acidification and salinity variations on the physiology of osmoregulating and osmoconforming crustaceans. J Comp Physiol B 188:729–738 https://doi.org/10.1007/s00360-018-1167-0
Raven J, Caldeira K, Elderfield H, Hoegh-Guldberg O and others (2005) Ocean acidification due to increasing atmospheric carbon dioxide. R Soc Policy Doc 12/05. Clyvedon Press, Cardiff.
Ries JB, Cohen AL, McCorkle DC (2009) Marine calcifiers exhibit mixed responses to CO2- induced ocean acidification. Geology 37:1131–1134 https://doi.org/10.1130/G30210A.1
Rosa R, Seibel BA (2008) Synergistic effects of climate-related variables suggest future physiological impairment in a top oceanic predator. Proc Natl Acad Sci 105:20776 – 20780
Sabine CL, Feely RA (2007) The oceanic sink for carbon dioxide. Pp. 31–49 in Greenhouse Gas Sinks. D. Reay, N. Hewitt, J. Grace, and K. Smith, eds, CABI Publishing, Oxfordshire, UK
Sabine CL, Feely RA, Gruber N, Key RM, Lee K, Bullister JL, Wanninkhof R, Wong CS, Wallace DW, Tilbrook B, Millero FJ, Peng TH, Kozyr A, Ono T, Rios AF (2004) The oceanic sink for anthropogenic CO2. Science 305(5682):367-71 doi: 10.1126/science.1097403
Seitz RD, Knick KE, Westphal M (2011) Diet selectivity of juvenile blue crabs (Callinectes sapidus) in Chesapeake Bay. Integr Comp Biol 51:598–607 https://doi.org/10.1093/icb/icr098
Small D, Calosi P, White D, Spicer JI, Widdicombe S (2010) Impact of medium-term exposure to CO2 enriched seawater on the physiological functions of the velvet swimming
32
crab Necora puber. Aquat Biol 10:11-21 https://doi.org/10.3354/ab00266
Spicer JI, Raffo A, Widdicombe S (2007) Influence of CO2-related seawater acidification on extracellular acid–base balance in the velvet swimming crab Necora puber. Mar Biol 151:1117–1125 https://doi.org/10.1007/s00227-006-0551-6
Stickle W, Wyler HJ, Dietz TH (2007) Effects of salinity on juvenile crab physiology and agonistic interactions between two species of juvenile blue crabs, Callinectes sapidus and Callinectes similis. J Exp Mar Biol Ecol 352:361–370 https://doi.org/10.1016/j.jembe.2007.08.013
Tans P, Keeling R (2017) Trends in atmospheric carbon dioxide. Annual mean growth rate of CO2 at Mauna Loa. NOAA Earth System Research Laboratory https://www.esrl.noaa.gov/gmd/ccgg/trends/gr.html
Taylor HH, Taylor EW (1992) Gills and lungs: The exchange of gases and ions. Pages 203-293 in F.W. Harrison and A.G. Humes (eds), Microscopic Anatomy of Invertebrates. Vol. 10. Decapod Crustacea. Wiley-Liss, New York.
Tomasetti SJ, Morrell BK, Merlo LR, Gobler CJ (2018) Individual and combined effects of low dissolved oxygen and low pH on survival of early stage larval blue crabs, Callinectes sapidus. PLoS ONE 13(12):e0208629 https://doi.org/10.1371/journal.pone.0208629
Towle DW, Palmer GE, Harris JL (1976) Role of gill Na+- K+‐dependent ATPase in acclimation of blue crabs (Callinectes sapidus) to low salinity. J Exp Zool Part A 196:315–321
Travis DF (1955a) The molting cycle of the spiny lobster, Panulirus argus Latreille. II. Pre- ecdysial histological and histochemical changes in the hepatopancreas and integumental tissues. Biol Bull 108:88-112
Van Colen C, Debusschere E, Braeckman U, Van Gansbeke D, Vincx M (2012) The Early Life History of the Clam Macoma balthica in a High CO2 World. PLoS ONE 7(9):e44655 https://doi.org/10.1371/journal.pone.0044655
CBNERR (2018) VECOS: Virginia Estuarine and Coastal Observing System. vecos.vims.edu. (Accessed 6 October 2020)
Vermeij GJ (1987) Evolution and escalation: An ecological history of life. Princeton University Press, Princeton, New Jersey.
Waldbusser GG, Bergschneider H, Green MA (2010) Size-dependent pH effect on calcification in post-larval hard clam Mercenaria spp. Mar Ecol Prog Ser 417:171–182 https://doi.org/10.3354/meps08809
Waldbusser GG, Voigt EP, Bergschneider H, Green MA, Newell RIE (2011) Biocalcification in the Eastern Oyster (Crassostrea virginica) in Relation to Long-term Trends in Chesapeake Bay pH. Estuaries Coast 34:221–231 https://doi.org/10.1007/s12237-010- 9307-0
Wang Y, Hu M, Wu F, Storch D, Pörtner H-O (2018) Elevated pCO2 Affects Feeding Behavior and Acute Physiological Response of the Brown Crab Cancer pagurus. Front Physiol
33
9:1164 doi: 10.3389/fphys.2018.01164
Welladsen HM, Southgate PC, Heimann K (2010) The effects of exposure to near-future levels of ocean acidification on shell characteristics of Pinctada fucata (Bivalvia: Pteriidae). Mollusc Res 30:125-130
Whiteley NM (2011) Physiological and ecological responses of crustaceans to ocean acidification. Mar Ecol Prog Ser 430:257–271 https://doi.org/10.3354/meps09185
Whiteley NM, Scott JL, Breeze SJ, Cann LMC (2001) Effects of water salinity on acid – base balance in decapod crustaceans. J Exp Biol 204(Pt 5):1003-11
Whiteley NM, Suckling CC, Ciotti BJ, Brown J, McCarthy ID, Gimenez L, Hauton C (2018) Sensitivity to near-future CO2 conditions in marine crabs depends on their compensatory capacities for salinity change. Sci Rep 8:15639 https://doi.org/10.1038/s41598-018- 34089-0
Wood H, Spicer JI, Widdicombe S (2008) Ocean acidification may increase calcification rates, but at a cost. Proc Royal Society B 275:1767–1773 doi: 10.1098/rspb.2008.0343
Wu F, Wang T, Cui S, Xie Z, Dupont S, Zeng J, Gu H, Kong H, Hu M, Lu W, Wang Y (2017) Effects of seawater pH and temperature on foraging behavior of the Japanese stone crab Charybdis japonica. Mar Pollut Bull 120(1-2):99-108 doi: 10.1016/j.marpolbul.2017.04.053
34
TABLES AND FIGURES Table 1. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab carapace strength (2018). k = number of parameters, including variance (σ2) as a parameter. Size = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH.
Variables
푥1 푥2 푥3 푥4 푥1푥3 푥1푥4 푥2푥3 푥2푥4 푥3푥4 Model k intercept Size ET Sal pH Size * Size * ET * ET * Sal * Sal pH Sal pH pH
g1 2 훽0
g2 3 훽0 β3
g3 3 훽0 β4
g4 4 훽0 β3 β4
g5 4 훽0 β1 β3
g6 4 훽0 β1 β4
g7 4 훽0 β1 β2
g8 4 훽0 β2 β3
g9 4 훽0 β2 β4
g10 5 훽0 β3 β4 β9
g11 5 훽0 β1 β2 β3
g12 5 훽0 β1 β2 β4
g13 5 훽0 β2 β3 β7
g14 5 훽0 β2 β4 β8
g15 5 훽0 β1 β4 β6
g16 5 훽0 β1 β2 β3 β4 β5 β6 β7 β8 β9
35
Table 2. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab claw pinch force (2018). k = number of parameters, including variance (σ2) as a parameter. Size = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH.
Variables
푥1 푥2 푥3 푥4 푥1푥3 푥1푥4 푥2푥3 푥2푥4 푥3푥4 Model k intercept Size ET Sal pH Size * Size * ET * ET * Sal * Sal pH Sal pH pH
g1 2 훽0
g2 3 훽0 β1
g3 3 훽0 β3
g4 3 훽0 β4
g5 4 훽0 β3 β4
g6 4 훽0 β1 β3
g7 4 훽0 β1 β4
g8 4 훽0 β1 β2
g9 4 훽0 β2 β3
g10 4 훽0 β2 β4
g11 5 훽0 β3 β4
g12 5 훽0 β1 β2 β3
g13 5 훽0 β1 β2 β4
g14 5 훽0 β2 β3 β7
g15 5 훽0 β2 β4 β8
g16 5 훽0 β1 β4 β6
g17 5 훽0 β1 β3 β5
g18 11 훽0 β1 β2 β3 β4 β5 β6 β7 β8 β9
36
Table 3. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab carapace strength (2019). k = number of parameters, including variance (σ2) as a parameter. SZ = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH.
Variables
푥1 푥2 푥3 푥4 푥5 푥3푥4 푥3푥5 푥4푥5 Model k intercept Size ET Sal pH Sex Sal * Sal * pH * pH Sex Sex
g1 2 훽0
g2 3 훽0 β4
g3 3 훽0 β3
g4 4 훽0 β3 β4
g5 4 훽0 β1 β4
g6 5 훽0 β1 β2 β4
g7 5 훽0 β3 β4 β6
g8 6 훽0 β1 β2 β3 β4
g9 6 훽0 β1 β3 β4 β6
g10 6 훽0 β2 β3 β4 β6
g11 7 훽0 β1 β2 β3 β4 β6
g12 7 훽0 β1 β2 β3 β4 β5
g13 8 훽0 β1 β2 β3 β4 β5 β6
g14 10 훽0 β1 β2 β3 β4 β5 β6 β7 β8
37
Table 4. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning blue crab claw pinch force (2019). k = number of parameters, including variance (σ2) as a parameter. SZ = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH.
Variables
푥1 푥2 푥3 푥4 푥5 푥3푥4 푥3푥5 푥4푥5 Model k intercept Size ET Sal pH Sex Sal * Sal * pH * pH Sex Sex
g1 2 훽0
g2 3 훽0 β1
g3 3 훽0 β4 β5
g4 4 훽0 β3 β5
g5 4 훽0 β3 β4 β5
g6 5 훽0 β1 β4 β5
g7 5 훽0 β1 β2 β4 β5
g8 6 훽0 β1 β3 β4 β5
g9 6 훽0 β3 β4 β5 β6
g10 6 훽0 β1 β2 β3 β4 β5
g11 7 훽0 β1 β3 β4 β5 β6
g12 7 훽0 β2 β3 β4 β5 β6
g13 8 훽0 β1 β2 β3 β4 β5 β6
g14 10 훽0 β1 β2 β3 β4 β5 β6 β7 β8
38
Table 5. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning clam growth and clam survival (2018 and 2019). k = number of parameters, including variance (σ2) as a parameter. Size = initial clam shell length. Sal = Salinity. pH = pH.
Variables
푥1 푥2 푥3 푥1푥2 푥1푥3 푥2푥3 Model k intercept Size pH Sal Size * Size * pH * pH Sal Sal
g1 2 훽0
g2 3 훽0 β1
g3 3 훽0 β2
g4 3 훽0 β3
g5 4 훽0 β1 β2
g6 4 훽0 β1 β3
g7 4 훽0 β2 β3
g8 5 훽0 β1 β2 β3
g9 5 훽0 β2 β3 β6
g10 5 훽0 β1 β2 β4
g11 5 훽0 β1 β3 β5
g12 6 훽0 β1 β2 β3 β6
g13 6 훽0 β1 β2 β3 β5
g14 6 훽0 β1 β2 β3 β4
g15 8 훽0 β1 β2 β3 β4 β5 β6
39
Table 6. Parameters for the multiple linear regression models corresponding to the different hypotheses concerning clam ridge rugosity (2018). k = number of parameters, including variance (σ2) as a parameter. Sal = Salinity. pH = pH.
Variables
푥1 푥2 푥1푥2 Model k intercept Sal pH Sal*pH
g1 훽 2 0
g2 3 훽0 β1
g3 3 훽0 β2
g4 4 훽0 β1 β2
g5 5 훽0 β1 β2 β3
40
Table 7. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab carapace strength (2018). k = number of parameters, including variance (σ2) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH. Model g4 had the highest wi = 0.651. Best model in bold.
Residual % deviance Adjusted Model Variables df AICc Δ w i i deviance explained R2
g1 NULL 2 111.6148 13.44551 < 0.01 1652.470238 0% 0.000
g2 Sal 3 114.829 16.65972 < 0.01 1641.309524 0.68% -0.076
g3 pH 3 111.0239 12.85463 < 0.01 1250.698611 24.31% 0.180
g4 Sal+ pH 4 114.9428 16.77359 < 0.01 1239.537897 25% 0.114
g5 SZ + SA 4 117.1223 18.9531 < 0.01 1448.33972 12.35% -0.036
g6 SZ + pH 4 111.9384 13.76915 < 0.01 1000.135672 39.48% 0.285
g7 SZ + ET 4 111.0652 12.896 < 0.01 939.6646706 43.14% 0.328
g8 ET + SA 4 114.2831 16.11383 < 0.01 1182.478637 28.44% 0.154
g9 ET + pH 4 106.8214 8.652195 0.013 693.9461716 58.01% 0.504
g10 Sal* pH 5 118.4491 20.27981 < 0.01 1109.680555 32.85% 0.127
g11 SZ+ET+SA 5 115.4 17.23075 < 0.01 892.5093511 45.99% 0.298
g12 SZ+ET+pH 5 98.16925 0 0.979 260.6736841 84.23% 0.795
g13 ET * SA 5 119.0822 20.91297 < 0.01 1161.018328 29.74% 0.087
g14 ET * pH 5 111.3939 13.22463 < 0.01 670.4075358 59.43% 0.473
g15 SZ * pH 5 116.985 18.81571 < 0.01 999.4932505 39.52% 0.214
g16 SZ * SA 5 120.4212 22.25194 < 0.01 1277.542294 22.69% -0.005
41
Table 8. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from multiple linear regression model g12 for blue crab carapace strength (2018). ET = Exposure time.
Parameter Variable Estimate SE 95% CI
Model g12
β0 Intercept 47.55916 10.49249 26.9942, 68.1241
1 Size 0.51071 0.12527 0.2652, 0.7562
2 ET -0.28006 0.05258 -0.3831, -0.1769
4 pH 15.75909 3.08779 9.7071, 21.8110
42
Table 9. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab carapace strength (2019). k = number of parameters, including variance (σ2) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 247.6762 8.897667 < 0.01 5831.594 0.00% 0
g2 pH 3 248.8538 10.07531 < 0.01 5584.104 4.24% 0.008241
g3 Sal 3 245.5966 6.818032 0.02 5009.56 14.10% 0.110282
g4 Sal+ pH 4 245.8441 7.065605 0.018 4619.884 20.78% 0.149101
g5 SZ + pH 4 251.1291 12.35055 < 0.01 5509.835 5.52% -0.01481
g6 SZ + ET + pH 5 253.998 15.21945 < 0.01 5504.126 5.62% -0.05275
g7 Sal*pH 5 238.7785 0 0.614 3314.089 43.17% 0.366128
g8 SZ + ET + Sal+ pH 6 250.7106 11.93208 < 0.01 4440.853 23.85% 0.116641
g9 SZ + Sal*pH 6 241.3872 2.608686 0.167 3254.591 44.19% 0.352608
g10 ET + Sal*pH 6 241.9211 3.142576 0.128 3313.029 43.19% 0.340984
g11 SZ + ET + Sal 7 244.8103 6.031774 0.03 3252.894 44.22% 0.325985
SZ + ET + Sal + pH g 7 253.9626 15.1841 < 0.01 4413.299 24.32% 0.085544 12 + Sex
SZ + ET + Sex + g 8 247.7041 8.925518 < 0.01 3159.652 45.82% 0.31684 13 Sal*pH
SZ + ET + Sal*pH + g 10 251.4864 12.70784 < 0.01 2679.989 54.04% 0.365364 14 Sal*Sex + pH*Sex
43
Table 10. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from multiple linear regression models for blue crab carapace strength (2019).
Parameter Variable Estimate SE 95% CI
Model g7
β0 Intercept 85.208 3.992 77.3848, 93.0318
3 Sal -1.637 5.843 -13.0893, 9.8155
4 pH -21.542 6.097 -33.4922, -9.5911
6 Sal*pH 26.692 8.340 10.3471, 43.0378
Model g9
β0 Intercept 100.2218 22.5713 55.9829, 144.4608
1 Size -0.1917 0.2835 -0.7473, 0.3639
3 Sal -0.9459 5.9929 -12.6919, 10.8000
4 pH -22.1334 6.2239 -34.3319, -9.9349
6 Sal*pH 25.9600 8.4975 9.3053, 42.6147
Model g10
β0 Intercept 84.591740 8.005736 68.9008, 100.2827
2 Exposure time 0.009826 0.109864 -0.2055, 0.2251
3 Sal -1.504602 6.138802 -13.5364, 10.5272
4 pH -21.473702 6.263341 -33.7496, -9.1978
6 Sal*pH 26.521399 8.715858 9.4386, 43.6042
44
Table 11. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab claw pinch force (2018). k = number of parameters, including 2 variance (σ ) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = crab size (carapace width). ET = Exposure time. Sal= Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 76.73979 2.011364 0.1124 269.7694 0% 0
g2 SZ 3 74.72843 0 0.3073 168.073 37.70% 0.314673
g3 Sal 3 79.43492 4.706496 0.0292 248.7893 7.78% -0.01445
g4 pH 3 76.1082 1.379771 0.1541 188.553 30.11% 0.231164
g5 Sal+ pH 4 79.40695 4.678528 0.0296 167.5729 37.88% 0.240791
g6 SZ + Sal 4 78.94333 4.214906 0.0374 161.2222 40.24% 0.269564
g7 SZ + pH 4 75.77909 1.050664 0.1817 123.8531 54.09% 0.438869
g8 SZ + ET 4 78.24757 3.51914 0.0529 152.1403 43.60% 0.310711
g9 ET + Sal 4 83.43346 8.705037 < 0.01 234.3841 13.12% -0.0619
g10 ET + pH 4 78.68624 3.957817 0.0425 157.8049 41.50% 0.285046
g11 Sal*pH 5 85.31513 10.5867 < 0.01 162.3829 39.81% 0.172343
g12 SZ + ET + Sal 5 84.15308 9.424657 < 0.01 147.3955 45.36% 0.248733
g13 SZ + ET + pH 5 79.34912 4.620698 0.0305 98.76961 63.39% 0.496577
g14 ET*Sal 5 89.37565 14.64723 < 0.01 227.7694 15.57% -0.16093
g15 ET*pH 5 84.758 10.02958 < 0.01 155.0162 42.54% 0.209891
g16 SZ*pH 5 81.59739 6.86896 < 0.01 119.1216 55.84% 0.392844
g17 SZ*Sal 5 84.82168 10.09326 < 0.01 155.841 42.23% 0.205687
45
Table 12. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for blue crab claw pinch force (2018). ET = Exposure time.
Parameter Variable Estimate SE 95% CI
Model g2
β0 Intercept -12.7849 10.00 -32.3891, 6.8193
1 Size 0.2839 0.1154 0.0577, 0.5100
Model g4
β0 Intercept 13.486 1.535 10.4768, 16.4948
4 pH -5.519 2.659 -10.7304, -0.3067
Model g7
β0 Intercept -7.1095 9.5885 -25.9027, 11.6836
1 Size 0.2342 0.1080 0.0225, 0.4459
4 pH -4.2128 2.3502 -8.8191, 0.3934
46
Table 13. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about blue crab claw pinch force (2019). k = number of parameters, including 2 variance (σ ) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = crab size (carapace width). ET = Exposure time. Sal = Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 144.0357 2.539928 0.085 727.576 0% 0
g2 SZ 3 142.4611 0.965252 0.186 599.0441 17.67% 0.135491
g3 pH + Sex 4 143.327 1.831182 0.121 543.18 25.34% 0.174853
g4 Sal+ Sex 4 148.2434 6.747595 0.010 679.199 6.65% -0.03177
g5 Sal+ pH + Sex 5 146.3972 4.901381 0.026 535.1695 26.44% 0.141857
g6 SZ + pH + Sex 5 141.4958 0 0.301 428.287 41.14% 0.313243
g7 SZ + ET + pH + Sex 6 145.2417 3.745903 0.046 426.2652 41.41% 0.276278
g8 SZ + Sal+ pH + Sex 6 144.2402 2.744382 0.076 407.2951 44.02% 0.308486
g9 Sex + Sal*pH 6 147.8535 6.357727 0.013 479.9976 34.03% 0.18505
SZ + ET + Sal + pH + g 10 Sex 7 148.3457 6.849863 0.010 401.8789 44.76% 0.275036
g11 SZ + Sex + Sal*pH 7 144.4231 2.92729 0.070 336.249 53.79% 0.393428
g12 ET + Sex + Sal* pH 7 146.6487 5.152892 0.023 372.0454 48.87% 0.328854
SZ + ET + Sex + g 13 Sal*pH 8 145.8828 4.38695 0.034 285.2686 60.79% 0.451087
SZ + ET + Sal*pH + g 14 Sal*Sex + pH*Sex 10 153.3859 11.89005 < 0.01 222.977 69.35% 0.50494
47
Table 14. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for blue crab claw pinch force (2019).
Parameter Variable Estimate SE 95% CI
Model g6
β0 Intercept -11.0459 11.4852 -33.5565, 11.4648
1 Size 0.2909 0.1324 0.0314, 0.5504
4 pH -3.6504 2.2986 -8.1556, 0.8548
5 Sex (Male) 5.1465 2.3229 0.5937, 9.6993
48
Table 15. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam growth (2018). k = number of parameters, including variance (σ2) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = Initial shell length. Sal = Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 -442.07 30.9123 < 0.01 0.01254 0.00% 0
g2 SZ 3 -440.58 32.4051 < 0.01 0.01243 0.89% -0.0045
g3 pH 3 -472.98 0 0.254 0.00812 35.29% 0.34417
g4 Sal 3 -441.06 31.9244 < 0.01 0.01235 1.51% 0.0018
g5 SZ + pH 4 -471.58 1.40738 0.126 0.00803 35.99% 0.34235
g6 SZ + Sal 4 -439.67 33.314 < 0.01 0.01222 2.59% -0.0007
g7 pH + Sal 4 -472.55 0.43469 0.205 0.00793 36.80% 0.35071
g8 SZ + pH + Sal 5 -471.31 1.67222 0.110 0.00782 37.67% 0.35078
g9 pH*Sal 5 -470.26 2.72845 0.065 0.00793 36.80% 0.34169
g10 SZ*pH 5 -470.41 2.57602 0.070 0.00791 36.93% 0.34301
g11 SZ*Sal 5 -437.76 35.2258 < 0.01 0.01216 3.08% -0.0096
g12 SZ + pH*Sal 6 -468.95 4.03017 0.034 0.00782 37.68% 0.34165
g13 pH + SZ*Sal 6 -470.11 2.87505 0.060 0.0077 38.62% 0.35158
g14 Sal + SZ*pH 6 -470.26 2.72015 0.065 0.00768 38.74% 0.3529
SZ*pH + SZ*Sal + g 8 -466.65 6.32866 0.011 0.00755 39.80% 0.34564 15 Sal*pH
49
Table 16. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for clam growth (2018). pH = pH. Sal = Salinity. SZ = Initial shell length.
Parameter Variable Estimate SE 95% CI
Model g3
β0 Intercept -0.0005263 0.0016989 -0.0040, 0.0028
2 pH -0.0152632 0.0024026 -0.0200, -0.0106
Model g5
β0 Intercept 0.0094149 0.0112822 -0.0127, 0.0315
1 Size -0.0007700 0.0008639 -0.0025, 0.0009
2 pH -0.0152247 0.0024063 -0.0199, -0.0105
Model g7
β0 Intercept -0.002105 0.002070 -0.0062, 0.0020
2 pH -0.015263 0.002391 -0.0199, -0.0106
3 Sal 0.003158 0.002391 -0.0015, 0.0078
Model g8
β0 Intercept 0.0089583 0.0112144 -0.0130, 0.0309
1 Size -0.0008643 0.0008610 -0.0026, 0.0008
2 pH -0.0152199 0.0023908 -0.0199, -0.0105
3 Sal 0.0033467 0.0023978 -0.0014, 0.0080
50
Table 17. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam growth (2019). k = number of parameters, including variance (σ2) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = Initial shell length. Sal = Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 -214.56 41.9848 < 0.01 0.18011 0.00% 0
g2 SZ 3 -217 39.543 < 0.01 0.16859 6.40% 0.05019
g3 pH 3 -229.8 26.7407 < 0.01 0.14041 22.04% 0.20894
g4 Sal 3 -232.82 23.7235 < 0.01 0.13449 25.33% 0.24231
g5 SZ + pH 4 -232.01 24.5328 < 0.01 0.13175 26.85% 0.24669
g6 SZ + Sal 4 -235.89 20.6542 < 0.01 0.12464 30.80% 0.28729
g7 pH + Sal 4 -253.37 3.16772 0.052 0.09709 46.09% 0.44484
g8 SZ + pH + Sal 5 -256.54 0 0.254 0.08977 50.16% 0.47893
g9 pH*Sal 5 -252.5 4.03882 0.034 0.0951 47.20% 0.44799
g10 SZ*pH 5 -231.09 25.4575 < 0.01 0.12914 28.30% 0.25039
g11 SZ*Sal 5 -235.14 21.3993 < 0.01 0.12187 32.34% 0.29261
g12 SZ + pH*Sal 6 -255.63 0.91321 0.161 0.08789 51.20% 0.48201
g13 pH + SZ*Sal 6 -256.4 0.14093 0.236 0.08692 51.74% 0.48769
g14 Sal + SZ*pH 6 -256.1 0.44672 0.203 0.0873 51.53% 0.48545
SZ*pH + SZ*Sal + g 8 -253.67 2.87265 0.060 0.08412 53.30% 0.48848 15 Sal*pH
51
Table 18. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for clam growth (2019). Size = Initial shell length. pH = pH. Sal = Salinity.
Parameter Variable Estimate SE 95% CI
Model g8
β0 Intercept 0.321333 0.124258 0.0778, 0.5649
1 Size -0.024677 0.010635 -0.0455, -0.0038
2 pH (low) -0.044793 0.008846 -0.0621, -0.0275
3 Sal (low) -0.049048 0.008829 -0.0664, -0.0317
52
Table 19. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam survival (2018). k = number of parameters, including variance (σ2) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = Initial shell length. Sal = Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 131.719 3.95424 0.041 20.4911 0% 0
g2 SZ 3 132.696 4.93138 0.025 20.2845 1.01% 0.00108
g3 pH 3 127.765 0 0.298 19.4107 5.27% 0.04411
g4 Sal 3 133.391 5.62629 0.018 20.4107 0.39% -0.0051
g5 SZ + pH 4 128.865 1.10033 0.172 19.2294 6.16% 0.04435
g6 SZ + Sal 4 134.397 6.63229 0.011 20.203 1.41% -0.004
g7 pH + Sal 4 129.452 1.68699 0.128 19.3304 5.66% 0.03934
g8 SZ + pH + Sal 5 130.582 2.81746 0.073 19.148 6.55% 0.03959
g9 pH*Sal 5 131.592 3.82745 0.044 19.3214 5.71% 0.03089
g10 SZ*pH 5 130.731 2.96572 0.068 19.1733 6.43% 0.03832
g11 SZ*Sal 5 135.094 7.32861 < 0.01 19.935 2.71% 0.00012
g12 SZ + pH*Sal 6 132.769 5.00362 0.024 19.1398 6.59% 0.03103
g13 pH + SZ*Sal 6 131.173 3.40783 0.054 18.869 7.92% 0.04473
g14 Sal + SZ*pH 6 132.481 4.7159 0.028 19.0907 6.83% 0.03351
SZ*pH + SZ*Sal+ g 8 135.39 7.6252 < 0.01 18.805 0.02984 15 Sal*pH 8.23%
53
Table 20. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for clam survival (2018). Size = Initial shell length. pH = pH. Sal = Salinity.
Parameter Variable Estimate SE 95% CI
Model g3
β0 Intercept 0.85714 0.05613 0.7471, 0.9672
2 pH -0.19643 0.07939 -0.3520, -0.0408
Model g5
β0 Intercept 0.55502 0.30323 -0.0393, 1.1493
1 Size 0.02407 0.02374 -0.0225, 0.0706
2 pH -0.19419 0.07941 -0.3498, -0.0386
Model g7
β0 Intercept 0.88393 0.06892 0.7488, 1.0190
2 pH -0.19643 0.07958 -0.3524, -0.0404
3 Sal -0.05357 0.07958 -0.2096, 0.1024
54
Table 21. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam survival (2019). k = number of parameters, including variance (σ2) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. SZ = Initial shell length. Sal = Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 147.912 16.9811 < 0.01 23.6786 0.00% 0
g2 SZ 3 147.383 16.4518 < 0.01 23.1267 2.33% 0.01443
g3 pH 3 149.855 18.9241 < 0.01 23.6429 0.15% -0.0076
g4 Sal 3 131.713 0.7817 0.204 20.1071 15.08% 0.14311
g5 SZ + pH 4 149.373 18.442 < 0.01 23.0934 2.47% 0.00682
g6 SZ + Sal 4 130.931 0 0.301 19.5874 17.28% 0.1576
g7 pH + Sal 4 133.665 2.7342 0.077 20.0714 15.23% 0.13678
g8 SZ + pH + Sal 5 132.932 2.00119 0.111 19.554 17.42% 0.15125
g9 pH*Sal 5 135.058 4.1264 0.038 19.9286 15.84% 0.13499
g10 SZ*pH 5 150.71 19.7787 < 0.01 22.9176 3.21% 0.00525
g11 SZ*Sal 5 133.072 2.1405 0.103 19.5783 17.32% 0.15019
g12 SZ + pH*Sal 6 134.381 3.44991 0.054 19.4174 18.00% 0.1493
g13 pH + SZ*Sal 6 135.117 4.18555 0.037 19.5453 17.46% 0.1437
g14 Sal + SZ*pH 6 134.054 3.12303 0.063 19.3608 18.23% 0.15178
SZ*pH + SZ*Sal + g 8 137.429 6.49801 0.012 19.1505 19.12% 0.14501 15 Sal*pH
55
Table 22. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for clam survival (2019). Size = Initial shell length. pH = pH. Sal = Salinity.
Parameter Variable Estimate SE 95% CI
Model g4
β0 Intercept 0.87500 0.05713 0.7630, 0.9870
3 Sal (low) -0.35714 0.08080 -0.5155, -0.1988
Model g6
β0 Intercept -0.40937 0.75731 -1.8937, 1.0750
1 Size 0.11101 0.06527 -0.0169, 0.2389
3 Sal -0.35556 0.08012 -0.5126, -0.1985
Model g8
β0 Intercept -0.38927 0.76161 -1.8820, 1.1035
1 Size 0.11077 0.06552 -0.0177, 0.2392
2 pH -0.03453 0.08042 -0.1921, 0.1231
3 Sal -0.35556 0.08042 -0.5132, -0.1979
Model g11
β0 Intercept -0.22912 1.10910 -2.4029, 1.9447
1 Size 0.09543 0.09574 -0.0922, 0.2831
3 Sal -0.69478 1.52120 -3.6763, 2.2867
5 Size*Sal 0.02934 0.13137 -0.2282, 0.2868
56
Table 23. AIC calculations for multiple linear regression models (gi) corresponding to hypotheses about clam ridge rugosity (2018). k = number of parameters, including variance 2 (σ ) as a parameter. AICc = corrected AIC value. ∆i = difference between model i and the best model in the set. wi = model probability of fitting the observed data. Sal = Salinity. pH = pH. Best model in bold.
Residual % deviance Adjusted Model Variables k AICc Δ w i i deviance explained R2
g1 NULL 2 47.1181 7.59433 0.015 8.27266 0.00% 0
g2 Sal 3 49.1634 9.63956 < 0.01 8.07401 2.40% -0.0204
g3 pH 3 39.5238 0 0.680 5.40325 34.69% 0.31717
g4 Sal + pH 4 41.5301 2.00628 0.249 5.20461 37.09% 0.31095
g5 Sal*pH 5 44.7434 5.21962 0.050 5.20141 37.13% 0.27694
57
Table 24. Estimate, SE (standard error), and 95% confidence interval (CI) of the parameters from the best multiple linear regression models for clam ridge rugosity (2018). pH = pH. Sal = Salinity.
Parameter Variable Estimate SE 95% CI
Model g3
β0 Intercept 1.2005 0.1431 0.9202, 1.4809
2 pH 0.6915 0.2023 0.2950, 1.0881
Model g4
β0 Intercept 1.1096 0.1760 0.7646, 1.4545
1 Sal 0.1820 0.2032 -0.2164, 0.5803
2 pH 0.6915 0.2032 0.2932, 1.0900
58
L-pH, H-sal L-pH, L-sal
H-pH, L-sal H-pH, H-sal 9 30 A 8.5 B 25 8
C)
° 7.5 20 7 15 pH 6.5 10 6 Temperature( 5 5.5 0 5 Week 1 Week 2 Week 11 Week 1 Week 2 Week 11
8 C 35 D 7 30
6 25 1) - 5 20 4 15
3 Salinity(ppt)
DO(mg L 10 2 5 1
0 0 Week 1 Week 2 Week 11 Week 1 Week 2 Week 11
Week Week
Figure 1 A-D. 2018 Water quality measurements over time in experimental treatments. DO = dissolved oxygen. In legend, L = low, H = High, sal = salinity. The high-pH, high-salinity treatment is the control. Error bars are standard error.
59
5000 A
4500
4000
3500
3000 L-pH, H-sal
2500 L-pH, L-sal H-pH, L-sal
2000 DIC DIC (umol) H-pH, H-sal 1500
1000
500
0 Week 1 Week 2 Week 11
5000 4500 B
4000
) 1
- 3500 3000 L-pH, H-sal 2500 L-pH, L-sal 2000 H-pH, L-sal
CA (umol CA kg SW 1500 H-pH, H-sal 1000 500 0 Week 1 Week 2 Week 11 Week
Figure 2 A-B. 2018 Water chemistry measurements over time in experimental treatments. DIC = dissolved inorganic carbon. CA = carbonate alkalinity. In legend, L = low, H = High, sal = salinity. The high-pH, high-salinity treatment is the control. Error bars are standard error.
60
8.5 A 8.5 B L-pH, H-sal L-pH, L-sal H-pH, L-sal H-pH, H-sal 8 8
7.5 7.5
pH 7 7
6.5 6.5
6 6 Week 2 Week 5 Week 10 Week 2 Week 5 Week 10
35 C 35 D 30 30 25 25 20 20
15 15 Salinity Salinity (ppt) 10 10 5 5 Week 2 Week 5 Week 10 Week 2 Week 5 Week 10
7 E 7 F
6.5 6.5
)
1 - 6 6
5.5 5.5 DO(mg L 5 5
4.5 4.5 Week 2 Week 5 Week 10 Week 2 Week 5 Week 10
30 G 30 H
C) 28 28 °
26 26
24 24
22 22 Temperature( 20 20 Week 2 Week 5 Week 10 Week 2 Week 5 Week 10 Week Week
Figure 3 A-H. Panels on the left are 2019 water quality measurements for clam tanks; panels on the right are 2019 water quality measurements for crab tanks. In legend, L = low, H = High, sal = salinity. The high-pH, high-salinity treatment is the control. Error bars are standard error.
61
L-pH, H-sal L-pH, L-sal H-pH, L-sal H-pH, H-sal
3500 A 3500 B
3000 3000
2500 2500
2000 2000
1500 1500
DIC DIC (umol) 1000 1000
500 500
0 0 Week 8 Week 10 Week 8 Week 10
3500 C 3500 D 3000 3000
) 1 - 2500 2500
2000 2000
1500 1500
1000 1000 CA CA (umolkgSW 500 500 0 0 Week 8 Week 10 Week 8 Week 10 Week Week
Figure 4 A-D. 2019 Water chemistry measurements over time in experimental treatments. DIC = dissolved inorganic carbon. CA = carbonate alkalinity. Panels on the left are 2019 water quality measurements for clam tanks; panels on the right are 2019 water quality measurements for crab tanks. In legend, L = low, H = High, sal = salinity. The high-pH, high-salinity treatment is the control. Error bars are standard error.
62
Figure 5. 2018 blue crab carapace strength in Newtons (N) by size in differing pH treatments.
63
Figure 6. 2018 blue crab carapace strength in Newtons (N) by exposure time in differing pH treatments.
64
L-pH, H-sal L-pH, L-sal H-pH, L-sal H-pH, H-sal
Figure 7. 2019 mean (± SE) adjusted carapace strength by treatment. Red indicates low-pH treatments and blue indicates high-pH treatments. The high-pH, high-salinity treatment is the control.
65
Figure 8. 2018 blue crab claw pinch force in Newtons (N) by carapace width. Data are color- coordinated by pH.
66
Figure 9. 2019 blue crab claw pinch force in Newtons (N) by carapace width. Data are color- coordinated by pH.
67
Figure 10. Box-and-whisker plot of 2019 blue crab claw pinch force in Newtons (N) by sex (F = female [n = 9], M = male [n = 13]). Dark horizontal lines are the median values, the bars indicate minimum and maximum data points, and circles are outliers.
68
A
L-pH, H-sal L-pH, L-sal H-pH, L-sal H-pH, H-sal
B Mean(slopes) growth clam
L-pH, H-sal L-pH, L-sal H-pH, L-sal H-pH, H-sal
Treatment
Figure 11 A-B. 2018 (panel A) and 2019 (panel B) box-and-whisker plots showing slopes of clam growth (from beginning to end of experiment) by treatment. N = 19 clams per treatment in both years. L = low, H = High, sal = salinity. Dashed lines mark 0. Dark horizontal lines are the median values, the bars indicate minimum and maximum data points, and circles are outliers. The high-pH, high-salinity treatment is the control.
69
A
B Survival (%) Survival
Treatment
Figure 12 A-B. 2018 (panel A) and 2019 (panel B) clam survival by treatment. L = low, H = High, sal = salinity. The high-pH, high-salinity treatment is the control.
70
A B
C D
Figure 13. Sample SEM images of the exterior of clam shells from different treatments. Panel A: shell from the control treatment (high-pH, high-salinity) with damage index score (DI) = 0. Panel B: shell from the low-pH, high-salinity treatment with DI = 1. Panel C: shell from the low-pH, low- salinity treatment with DI = 3. Panel D: shell from the high-pH, low-salinity treatment with DI = 2. Shell deterioration circled in red.
71
2.5 (1.5) (2) (1.17) (0.5)
2
1.5 Rugosity 1
0.5
0 L-pH, H-sal L-pH, L-sal H-pH, L-sal H-pH, H-sal Treatment
Figure 14. Mean clam ridge rugosity (+ SE), or deterioration, by treatment. Damage Index (DI) scores bolded in parentheses above corresponding treatment. L = low, H = High, sal = salinity. The high-pH, high-salinity treatment is the control.
72
100 A
80
60 Time eating (%) 40 Time moving (%)
Time spent(%) Time Time buried (%) 20
0 High (n = 10) Low (n = 4) Salinity
B 100
80
60 Time eating (%)
40 Time moving (%)
spent(%) Time Time buried (%) 20
0 High (n = 6) Low (n = 8) pH
Figure 15 A-B. Stacked bar graphs of crab behavior in percentages by (A) salinity and (B) pH. The number of trials for each parameter (high/low salinity, high/low pH) is indicated in parentheses.
73