Multiple Physiological Responses to Multiple Environmental Challenges: An Individual Approach

Integrative and Comparative Biology Advance Access published May 9, 2013

Integrative and Comparative Biology Integrative and Comparative Biology, pp. 1–11 doi:10.1093/icb/ict041

Society for Integrative and Comparative Biology

SYMPOSIUM

Multiple Physiological Responses to Multiple Environmental Challenges: An Individual Approach P. Calosi,1 L. M. Turner, M. Hawkins, C. Bertolini, G. Nightingale, M. Truebano and J. I. Spicer Marine Biology and Ecology Research Centre, School of Marine Science and Engineering, Plymouth University, Drake Circus, Plymouth, Devon PL4 8AA, UK From the symposium ‘‘Physiological Responses to Simultaneous Shifts in Multiple Environmental Stressors: Relevance in a Changing World’’ presented at the annual meeting of the Society for Integrative and Comparative Biology, January 3–7, 2013 at San Francisco, California. E-mail: [email protected]

Synopsis The injection of anthropogenically-produced CO2 into the atmosphere will lead to an increase in temperature and a decrease in pH at the surface of the oceans by 2100. Marine intertidal organisms possess the ability to cope in the short term with environmental fluctuations exceeding predicted values. However, how they will cope with chronic exposure to elevated temperature and pCO2 is virtually unknown. In addition, individuals from the same species/population often show remarkable levels of variation in their responses to complex climatic changes: in particular, variation in metabolic rates often is linked to differences in individuals’ performances and fitness. Despite its ecological and evolutionary importance, inter-individual variation has rarely been investigated within the context of climatic changes, and most investigations have typically employed orthogonal experimental designs paired to analyses of independent samples. Although this is undoubtedly a powerful and useful approach, it may not be the most appropriate for understanding all alterations of biological functions in response to environmental changes. An individual approach arguably should be favored when trying to describe organisms’ responses to climatic change. Consequently, to test which approach had the greater power to discriminate the intensity and direction of an organism’s response to complex climatic changes, we investigated the extracellular osmo/iono-regulatory abilities, upper thermal tolerances (UTTs), and metabolic rates of individual adults of an intertidal amphipod, Echinogammarus marinus, exposed for 15 days to combined elevated temperature and pCO2. The individual approach led to stronger and different predictions on how ectotherms will likely respond to ongoing complex climatic change, compared with the independent approaches. Consequently, this may call into question the relevance, or even the validity, of some of the predictions made to date. Finally, we argue that treating individual differences as biologically meaningful can lead to a better understanding of the physiological responses themselves and the selective processes that will occur with complex climatic changes; selection will likely play a crucial role in defining species’ responses to future environmental changes. Individuals with higher metabolic rates were also characterized by greater extracellular osmo/iono-regulative abilities and higher UTTs, and thus there appeared to be no evolutionary trade-offs between these functions. However, as individuals with greater metabolic rates also have greater costs for maintenance and repair, and likely a lower fraction of energy available for growth and reproduction, trade-offs between life-history and physiological performance may still arise.

Introduction It is estimated that the continuing increase in atmospheric CO2 from anthropogenic sources will lead to an increase in surface temperatures of the ocean by 3–58C (Sokolov et al. 2009), and a decrease in surface pH of the ocean by 0.3–0.5 units by 2100 (Caldeira and Wickett 2003). However, marine intertidal organisms already experience periodical

environmental fluctuations that exceed these values (Newell 1979; Morris and Taylor 1983; Agnew and Taylor 1986) and are considered to have already evolved the machinery necessary to cope with short-term exposure to extreme environmental conditions (Widdicombe and Spicer 2008; Melzner et al. 2009; Whiteley 2011). Our knowledge of their ability to cope with chronic exposure to elevated

ß The Author 2013. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved. For permissions please email: [email protected].

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

1

2

traits are often best described using an individual approach, rather than by ‘‘logical triangulation’’ of results emerging from analyses of independent sample. Emerging evolutionary and functional trade-offs determine how the physiological machinery of marine organisms operates, and the constraints some species could face in the near future with ongoing climatic changes (e.g., Po¨rtner et al. 2006). Therefore, the use of an individual-based approach arguably should be adopted when trying to detect and describe both intensity and direction of a variable organismal response to climatic change. In addition, the utilization of an individual-based approach directly places research into an evolutionary context, which has been advocated as a priority in the investigation of the effect of global changes (Kelly and Hofmann 2012). Consequently, the aim of this study is to compare an individual-based approach with an approach using orthogonal experimental designs paired to analysis of independent samples. We wanted to know which of the two approaches had the greater power to discriminate the intensity and direction of an organism’s response to complex climatic changes. To do so, metabolic rates, the extracellular osmo/iono-regulative ability (using Naþ/Kþ-ATPase as a proxy), and upper thermal tolerance (UTT) of adult individuals of an intertidal amphipod, Echinogammarus marinus Leach (1875), were investigated after 15 days exposure to combined elevated temperature and pCO2. Echinogammarus marinus encounters significant temporal and spatial variation in environmental factors where it occurs on the mid-shore and could be considered as physiologically hardy (Lincoln 1979). However, adult amphipods of various species have been shown to be potentially physiologically sensitive to the warming (e.g., Rastrick and Whiteley 2011) and acidification (e.g., Hauton et al. 2009) of the ocean. Data were analyzed using both an independent and an individual sample approaches, and results of both analyses formally compared and differences between them critically assessed.

Materials and methods Collection and husbandry of animals Echinogammarus marinus (mean wet mass  SE, 0.98  0.002 g) were collected by hand from the intertidal mudflats at Saltash, Cornwall, UK (508240 51.5700 N, 48120 41.700 W) in May 2012. All individuals were placed in a plastic bucket (volume ¼ 25 l) half-filled with damp weed (Ascophyllum nodosum) to prevent desiccation and mechanical damage

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

temperature and/or pCO2 is rather limited (e.g., Small et al. 2010; Thomsen and Melzner 2010; Melatunan et al. 2011; Hammer et al. 2012). Metabolic rate is considered the most fundamental of all biological rates (Brown et al. 2004). It largely defines the physiological and life-history performances of ectotherms (Prosser 1991; Brown et al. 2004), ultimately affecting their Darwinian fitness (Calow and Forbes 1998; Spicer and Gaston 1999; Crnokrak and Roff 2002; Brown et al. 2004; Burton et al. 2011). However, metabolic rates are highly variable among individuals, even from the same population (see Aldrich 1975; Spicer and Gaston 1999), with such variation often linked to differences in an individual’s performance and fitness (Aldrich 1975, 1989; Bennett 1987; Hayes et al. 1992; Spicer and Gaston 1999; Ve´zina et al. 2006). Within the context of warming and acidification of the ocean, inter-individual variation and its ecological implications rarely have been investigated (cf. for life-history traits, see Pistevos et al. 2011; Sunday et al. 2011; Foo et al. 2012; Schlegel et al. 2012), particularly with respect to metabolic rate. Most investigations of the potential effect of complex global climatic changes typically have employed orthogonal experimental designs paired to analyses of independent samples. Although these types of analysis are undoubtedly useful and powerful, they may not be appropriate for understanding all alterations of biological functions and traits in response to an environmental change. For example, it was long considered that exposure of crustaceans to environmental hypoxia resulted in an increase in the respiratory pigment hemocyanin (Hagerman 1986; DeFur et al. 1990), but the response was so variable that often the increase was not statistically significant (Senkbeil and Wriston 1980) and in some cases, even a decrease was observed (Baden et al. 1990); sometimes all these responses occurred within the same species, for example, the Norway lobster Nephrops norvegicus Linnaeus (1758) (Hagerman and Uglow 1985; Baden et al. 1990; Hagerman et al. 1990). However, by taking an inter-individual approach, Spicer and Baden (2001) showed that the magnitude of an hypoxia-induced change in hemocyanin concentration ([Hc]) varied in a predictable way from individual to individual; those with low initial [Hc] showed a marked increase, those with the highest [Hc] showed a marked decrease, and at a particular initial [Hc], there was no significant difference at all. Thus, the overall response observed was a function of the initial [Hc] of individuals used in the experiment. This illustrates that co-variation and trade-offs among biological

P. Calosi et al.

3

Global change and the individual approach

Experimental design and setup To investigate the effect of elevated temperature and pCO2 on some aspects of physiological function, an orthogonal experimental design incorporating two levels of seawater temperature and two levels of pCO2 was used. The two levels of pCO2 (380 and 1000 matm, pHNBS 8.05 and 7.70, respectively) corresponded to current and predicted (year 2100; Caldeira and Wickett 2003) global ocean mean pCO2 values, respectively. The two seawater temperature levels (158C and 208C) corresponded to the mean monthly sea surface temperature (SST) at the collection site (Joyce 2006) and to an increase of þ58C in line with predicted warming trends for SST (Sokolov et al. 2009). Amphipods were exposed for 20 days to one of four treatments: ‘‘control’’ (158C þ pCO2 380 matm), ‘‘elevated pCO2’’ (158C þ pCO2 1000 matm), ‘‘elevated temperature’’ (208C þ pCO2 380 matm), or ‘‘combined’’ (208C þ pCO2 1000 matm). Four replicate aquaria were utilized per treatment to avoid pseudoreplication. Amphipods were haphazardly divided into four groups of 120 individuals, each assigned to one of the four treatments described above. Each group was further sub-divided into four groups of 30 individuals, each haphazardly allocated to one of the four replicate aquaria. All aquaria were then transferred to the NERC-funded Experimental Gas and Temperature Manipulation Mesocosm (see below for description) held inside a controlled-temperature room kept at 158C with L:D 12:12 at MBERC, and individuals exposed to experimental conditions for 20 d. During this period, individuals were fed carrot ad libitum as previously described. Each aquarium was supplied

with 15 weighted hair rollers, to provide shelter and a complex substrate for adherence by amphipods. The Experimental Gas and Temperature Manipulation Mesocosm is a modified and integrated version of the equilibration flow-through systems used by Widdicombe et al. (2009) and Melatunan et al. (2011). Briefly, each treatment level within the system consisted of a header tank (volume ¼ 80 L) of sea water, supplied from one of two sumps (158C and 208C) and aerated with either air (pCO2 380 matm), or CO2enriched air (pCO2 1000 matm). A submersible pump (Rio 200 pump; American Aquarium Products, Grants Pass, OR, USA) was used to aid mixing in each header tank. CO2 gas was slowly released into a Buchner flask for mixing using a multistage CO2 regulator (EN ISO 7291; GCE, Worksop, UK). CO2 level in the CO2-enriched air was logged continuously using a CO2 analyser (280; LI-COR, Lincoln, NE, USA) and adjusted manually to the desired level on a daily basis. From each of the four header tanks, sea water was gravity-fed at a constant rate (60 mL min1) to one of the four replicate aquaria (transparent sealed 5 l containers), held within larger holding trays (volume ¼ 300 L), into which excess sea water was allowed to flow. The holding trays were further supplied with sea water from the sump at the appropriate temperature, effectively creating a water bath and maintaining the desired temperature in the aquaria. Sea water overflowed from the trays to a collective sump where it was filtered, aerated, and recirculated, via a submersible pump (1262; EHEIM GmbH and Co. KG, Deizisau, Germany) to the header tanks and trays. Approximately 10% of the sea water in the system was replaced weekly, and deionized water was added as needed to maintain stable salinity levels. Seawater temperature was increased to 208C in the elevated temperature treatment using aquarium heaters (50 W aquarium heater; EHEIM Jager GmbH and Co. KG, Stuttgart, Germany) placed in the header tanks and holding trays. At the end of the exposure period, metabolic rates (MO2) were determined for all individuals. The group within a treatment was then sub-divided with half of the amphipods used for analysis of gill Naþ/Kþ-ATPase activity, and the other half used to determine UTT. Determination of standard MO2 and wet mass Standard MO2 was determined using closed respirometers. MO2 was measured for individual amphipods using closed, blackened-out, glass incubation chambers (volume ¼ 23 mL). Each chamber was

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

during transport. Buckets were transferred to the experimental aquarium facility at the Marine Biology and Ecology Research Centre (MBERC) at Plymouth University (Plymouth, UK) 560 min after collection. Upon arrival, amphipods (only males were used in experiments) were exposed to constant conditions for 14 days to remove any effects of differences in recent environmental history. This was achieved by placing amphipods in a number of aquaria (volume ¼ 30 L), filled with aerated natural sea water T (8C) ¼ 15  1, S ¼ 33  1, pH ¼ 8.1  0.1 and total alkalinity (AT) 2213  1 mEq Kg1. All values are means  1 SE. Individuals were exposed to a 12 h:12 h L:D cycle and fed carrot ad libitum. Sea water was fully replaced every 2 d to maintain quality. Salinity, temperature, and pHNBS were measured daily.

4

Determination of gill Naþ/Kþ-ATPase activity Naþ/Kþ-ATPase activity was determined in gill tissue using a modification of the protocol of Brooks and Lloyd Mills (2003). Briefly, gills were carefully dissected from the amphipods and sonicated on ice (MicrosonTM XL ultrasonic cell disruptor; Qsonica, Newton, MA) in 250 mL sonication buffer at 48C (100 mmol L–1 HEPES, 100 mmol L–1 NaCl, and 0.1% sodium deoxycholate, pH 7.2). ATPase activity was determined by adding 30 ml of homogenate to 500 ml of each of two different buffers. The first buffer was constructed using 10 mmol L–1 MgCl2, 100 mmol L–1 NaCl, 15 mmol L– 1 KCl, and 100 mmol L–1 HEPES, adjusted to pH 7.2. The second buffer was constructed exactly the same way as the first buffer except that ouabain

(10 mmol L–1, which specifically inhibits Naþ/KþATPase activity) was substituted for KCl. Samples were prepared in triplicate. The reaction was initiated by the addition of 27 ml ATP (100 mmol L–1) as substrate and incubated at 378C for 20 min on a heat block (Dry Block Thermostat; Grant Instruments Cambridge Ltd, Shepreth, UK). The reaction was terminated by the addition of 1 ml Bonting’s reagent (560 mmol L–1 H2SO4, 8.1 mmol L–1 ammonium molybdate, and 176 mmol L–1 FeSO4). The resultant mixture was left to stand for 20 min at room temperature before absorbance was measured ( ¼ 700 nm) spectorphotometrically (Novaspec II, Pharmacia Biotech, Uppsala, Sweden) with 0.65 mmol l1 phosphorus standard solution (Sigma-Aldrich Company Ltd, Gillingham, UK) as a standard. The difference between ATP concentrations in the two buffers can be attributed to Naþ/Kþ-ATPase activity. Protein concentration was also determined for the gill homogenate in a microplate format (VersaMax microplate reader; Molecular Devices, Sunnyvale, CA) using the method of Bradford (1976) with 200 mg mL1 bovine serum albumin (Sigma-Aldrich Company Ltd) as the standard. Determination of Upper Thermal Tolerance To measure UTT, a dynamic thermal-limit analysis was conducted (Lutterschmidt and Hutchison 1997a, 1997b), using a number of observable responses identified during preliminary trials (as in Calosi et al. 2008; Massamba-N’Siala et al. 2012). A temporal sequence of responses to increasing temperature was identified as follows: (1) circular swimming (CS), continuous CS, (2) no tail flexing (NTF), major decrease in activity following CS accompanied by NTF, or contraction or expansion of the uropod region, (3) no pleopod beating (NPB) (pleopods stop beating; amphipods lay on the bottom of the experimental well for 415 s, and do not respond to gentle prodding); this end-point was considered ‘‘death,’’ based on preliminary trials showing that individuals did not recover once this stage was reached. Experiments to determine UTT were initiated at pre-exposure temperatures in each case. Temperature was gradually increased at a rate of 18C min1 (see Rezende et al. 2011) using a computer-controlled water bath (R5; Grant Instruments Cambridge Ltd). Amphipods were placed individually into one well (diameter 35 mm, depth 20 mm) of a six-well plate, the bottom of which was previously painted white to allow visualization of the amphipods and identification of end-points. A maximum of five individuals were examined at any time, and measurements were

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

supplied with pre-filtered (0.22 mm) sea water at the appropriate temperature and pCO2, and equipped with a meshed cylinder connecting top and bottom of the chamber. The inside of the cylinder contained a magnetic flea, while an individual amphipod was placed on the outside of the cylinder, to prevent contact between the amphipod and the magnetic flea as well as to restrict activity of the amphipod. Sealed while submerged, the chambers were placed onto a multi-channel magnetic stirrer (MIX 15 eco; 2mag AG, Munich, Germany) to ensure moderate mixing of water, and to prevent stratification of oxygen within the respirometer. Oxygen levels (mbar) in the chambers were measured every 5 min using a calibrated optical O2 analyzer (5250i, OxySense, Dallas, TX) in combination with an external probe (101, OxySense) and a fluorescent disc placed inside the chamber (Oxydot, OxySense). MO2 was measured over a 3 h period, or until O2 saturation reached 70%. All apparatus were located in CT-rooms at the appropriate temperature level. MO2 was expressed as nmol O2 min1 wet mass g1 salinity-temperature-pressure. Upon completion of measurements of MO2, amphipods were removed from the chamber using a tea strainer, gently blotted dry, and weighed (Electronic high-precision scale, PF-203; Fisher Scientific UK Ltd, Loughborough, UK). Within 10 s after measurement, individuals were placed into labeled perforated tubs (volume ¼ 10 mL) and into a 1 L aquarium containing sea water at the appropriate temperature and pCO2, then returned (after 53 min) to the aquaria from which they had been removed. Here, they were left to recover for approximately 1 h before either Naþ/Kþ-ATPase activities or UTT were measured.

P. Calosi et al.

5

Global change and the individual approach

all undertaken by the same observer (P.C.) assisted by a scribe (C.B.) in order to avoid observer bias (Terblanche et al. 2007). A transparent lid was used to prevent escapes and minimize evaporation. At every end-point, temperature was measured within the empty well with a digital thermometer (HH802U; OmegaÕ Engineering Inc., Stamford, CT, USA). Physico-chemical environment

Statistical analyses The relevance of employing an independent versus an individual-sample approach to interpret the effect of elevated temperature and pCO2 (in isolation and combined) on amphipods’ Naþ/Kþ-ATPase activity and UTT was examined. However, as our experimental design included four replicate aquaria per treatment, we ascertained in a preliminary analysis whether our use of ‘‘Aquaria’’ as a random factor nested within temperature  pCO2, had a significant effect on any of the parameters investigated. In all cases, ‘‘Aquaria’’ had no significant effect (P40.05), so this term was removed from further analyses. Power analysis performed on datasets analyzed by both approaches (independent and individual) revealed acceptable statistical power: observed power for the most complex significant statistical terms (i.e., single factors, two-way and thee-way interactions) ranged between 0.518 and 0.899 for the independent approach, and between 0.820 and 1.000 for the individual approach;  ¼ 0.05. Analysis of independent samples

A two-way analysis of covariance (ANCOVA) with temperature and pCO2 as fixed factors and wet mass as a covariate was used for comparisons of independent samples of Naþ/Kþ-ATPase activity and UTT values collected for the different treatments. When not significant, wet mass was removed from the analysis.

Table 1 Mean ( SE) values of the physico-chemical parameters of sea water during the experiment Source

Cold

Cold acidified a

Salinity

33.45  0.14

33.50  0.16

a

Temperature (8C)

16.12  0.09

8.07  0.02a

AT (mEq Kg ) DIC (mmol Kg1)* pCO2 (matm)* [HCO3] (mmol Kg1)* 2

1

[CO3 ] (mmol Kg )*

calcite*

aragonite*

a

2102  18

33.76  0.08a

b

19.05  0.06b

8.05  0.12a

7.79  0.01b

18.91  0.06

7.83  0.02b

a

1

Warm acidified a

33.80  0.08

a

16.24  0.11

pH

Warm a

2065  15

2225  9

b

2234  9b

1936.40  16.96a

1986.20  13.43b

a

479.60  56.10

b

868.40  47.80

544.70  28.40

1074.10  24.10c

1797.50  20.87a

1880.90  17.80b

1889.80  9.63b

2035.60  10.77c

a

121.51  5.11

b

73.63  4.34

2043.79  9.88c

2152.11  11.75c a

a

80.63  2.10b

135.71  2.60

2.93  0.12a

1.78  0.10b

3.27  0.06a

1.94  0.07b

a

b

a

1.26  0.04b

1.88  0.08

1.14  0.07

2.12  0.03

Seawater salinity, temperature, pH, alkalinity (AT), dissolved inorganic carbon (DIC), partial pressure of CO2 (pCO2), bicarbonate and carbonate concentrations ([HCO3], [CO32]), and saturation status of calcite and aragonite ( calcite and aragonite). Asterisks indicate calculated parameters. Different superscript letters indicate significant differences between treatments.

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Temperature, salinity, and pH of sea water taken from the 14 d acclimation aquaria and from experimental aquaria were measured daily. Salinity was measured using a refractometer (D&D The Aquarium Solution Ltd, Ilford, UK). Temperature was measured using digital thermometer (TL; Fisher Scientific, Loughborough, UK). pH was measured using a microelectrode (InLabÕ Expert Pro; Mettler-Toledo Ltd, Beaumont Leys, UK) coupled to a calibrated pH meter (Seven Easy; Mettler-Toledo Ltd, Beaumont Leys, UK). Total alkalinity (AT) was measured from sea water samples collected four times during the experimental period. Samples were transferred to 150 mL borosilicate bottles with Teflon caps, poisoned with 30 ml of saturated HgCl2 solution (0.02 % sample volume) and kept in the dark until measured by Gran titration (As-Alk2; Apollo SciTech Inc., Bogart, GA). Partial pressure of carbon dioxide (pCO2), concentrations of bicarbonate and carbonate ions ([HCO3] and [CO32]), saturation state of calcite and aragonite (Vcalcite and Varagonite), and dissolved inorganic carbon were calculated at the end of the experiment (Table 1), using CO2 SYS (Pierrot et al. 2006), employing constants from Mehrbach et al.

(1973) refitted to the NBS pH scale by Dickson and Millero (1987) and the KSO4 dissociation constant from Dickson (1990).

6 Analysis of individual samples

A ANCOVA with temperature and pCO2 as fixed factors with individual MO2 and wet mass as covariates was used for the comparisons of individual Naþ/Kþ-ATPase activity and UTT values collected for the different treatments. When not significant, wet mass was removed from the analysis. All analyses were carried out in SPSS 17. Data were approximately normally distributed and variances were homogeneous (P40.05).

Results Analysis of independent samples

Analyses of individual samples Results for the analyses of ‘‘individual samples’’ are shown in Fig. 1E–H. For all analyses of individual samples, the term ‘‘residual of individuals’ MO2’’ had the highest F value, indicating its importance in defining amphipod Naþ/Kþ-ATPase activity and UTL, relative to other terms. Based on the analyses of independent sample, Naþ/Kþ-ATPase activity was significantly affected

by the interaction between temperature and pCO2 but differently according to individuals’ MO2 levels (here as allometric residuals of individuals’ MO2) (Fig. 1E), as indicated by the presence of a significant three-way interaction between these factors (F1,148 ¼ 73.661, P50.0001). This model has a high predictive ability (R2 ¼ 0.574, R2adj ¼ 0.553). Similarly, UTT–NTF was significantly affected by the interaction between temperature and pCO2 but differently according to individuals’ MO2 levels (Fig. 1G), as indicated by the presence of a significant three-way interaction between these factors (F1,154 ¼ 10.104, P ¼ 0.002). Again the model shows a good predictive ability (R2 ¼ 0.593, R2adj ¼ 0.573). UTT–NPB was greater in individuals with higher MO2 (Fig. 2), as indicated by the presence of a positively significant relationship between UTT–NPB and individual MO2 levels (F1,153 ¼ 118.386, P50.0001). The model showed a high predictive ability (R2 ¼ 0.444, R2adj ¼ 0.432). No significant effect of temperature, pCO2, or their interaction was detected for UTT–CS (Fig. 1F, P40.05). The patterns of significance reported above are mirrored by AIC values (not reported here), in which the significant terms emerging from the ANOVA-type analysis within the General Linear Model test are always those with the lowest AIC values (here applying the concept ‘‘the lower AIC value the better model’’).

Discussion Here, we show that the utilization of an individual approach, as opposed to an independent approach, lead to stronger and different predictions on how ectotherms will respond to ongoing complex climatic changes; even when, as here, a quadratic experimental design is employed. Consequently, this may call into question the relevance, or even validity, of some predictions made on the effect of climatic changes on organismal function based on analyses of independent samples. Below we discuss our findings in more detail and highlight their potential to provide better predictions on the effects of global change on organismal function, when compared with more well-established approaches. We also argue that treating individual differences as biologically meaningful (and not just ‘‘noise’’) can lead to a greater understanding of physiological responses (as in Spicer and Baden 2001; Calosi et al. 2005) and improve our understanding of complex climatic changes (as for Melatunan et al. 2013).

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Results for the analyses of ‘‘independent samples’’ are presented in Fig. 1A–D. Mean Naþ/Kþ-ATPase activity was greater at the higher exposure temperature for the low pCO2 treatments, whereas no significant difference with temperature was found at the higher pCO2 treatments (Fig. 1A), as indicated by the presence of a significant interaction between temperature and pCO2 (F1,153 ¼ 6.550, P ¼ 0.011). However, this model has very low predictive ability (R2 ¼ 0.081, R2adj ¼ 0.063). Mean UTT, measured as ‘‘CS,’’ was lower at the lower pCO2 of exposure for the lower temperature of exposure, while exposure to different pCO2 levels had no significant effect at the higher temperature treatments (Fig. 1B), as indicated by the presence of a significant interaction between pCO2 and temperature (F1,153 ¼ 4.072, P ¼ 0.045). Wet mass had a significant positive effect on this parameter (F1,153 ¼ 4.243, P ¼ 0.041). The predictive ability of this model is also very low (R2 ¼ 0.090, R2adj ¼ 0.075). Mean UTT, measured as ‘‘NPB,’’ was lower at the lower temperature of exposure (Fig. 1D), as indicated by the presence of a significant positive effect of temperature on this parameter (F1,155 ¼ 10.588, P ¼ 0.001). The predictive ability of this model was also very low (R2 ¼ 0.073, R2adj ¼ 0.061). Finally, no significant effect of temperature, pCO2, or their interaction was detected on UTT measured as ‘‘NTF’’ (Fig. 1C, P40.05).

P. Calosi et al.

Global change and the individual approach

7

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Fig. 1 The effect of elevated temperature and pCO2 on Naþ/Kþ-ATPase activity and upper thermal tolerance of Echinogammarus marinus Leach (1875), measured as circular swimming, no tail flipping, and no pleopod beating. Results are presented in the standard format of an independent-sample approach (A–D), with histograms presenting mean  1 SE; and in relation to individuals’ MO2 values in the standard format of an individual-sample approach (E–H).

8

P. Calosi et al.

Why is it important to use an individual approach? Individual variation has been recognized as a fundamental property of biological systems even since Darwin (1859). Such variation often explains differences in individuals’ performance and fitness (Bennett 1987; Aldrich 1989; Hayes et al. 1992; Spicer and Gaston 1999; Ve´zina et al. 2006); thus, its importance and that of utilizing an individual approach have been advocated on multiple occasions (e.g., Bennett 1987; Aldrich 1989; Spicer and Gaston 1999). Despite this, researchers interested in how animals work have rarely consider an individual-sample approach as the preferred one in the investigation of organisms’ responses to multiple environmental challenges (although cf. Hayes et al. 1992; Pistevos et al. 2011). This stands in stark contrast to other areas of Biology such as ecotoxicology (e.g., Depledge 1994), community ecology (e.g., Hanski and Singer 2001, Bolnick et al. 2011), ecosystem function (e.g., Cianciaruso et al. 2009), and behavior (particularly the study of ‘‘personality,’’ e.g., Re´ale et al. 2010) where there has both been a resurgence and often a long-standing interest in addressing questions that are better analyzed using an individual approach. Exactly why comparative animal physiology

and its derivations should be so far behind in using an individual approach is not clear. This study shows that, generally, the inclusion of individual-level responses in the investigation of the potential effect of the combination of elevated temperature and pCO2 improves the overall predictability of statistical models. The average increase in R2 and R2adj is þ0.34 and þ0.39, respectively, across all variables investigated, and it reaches þ0.50 for both parameters when only statically significant models are considered, that is, Naþ/Kþ-ATPase activity and UTT measured as NTF had significant statistical models both from the independent and the individual-sample analyses. Alternatively, we note with some concern that the type and number of ‘‘terms’’ included in the statistical models emerging from the analyses, and their statistical significance, changed markedly when using an independent, as opposed to an individual analysis (Table 2). For instance, for UTT as NTF and NPB, no significant effect of the factors investigated was found for the independent analysis whereas it was found for the individual analysis, and for UTT CS, the opposite pattern was observed. Furthermore, individual MO2 appears as a major factor driving the responses of E. marinus to exposure to elevated

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Fig. 2 The relationship between residuals for individuals’ MO2 and residuals for individuals’ (A) Naþ/Kþ-ATPase, (B) upper thermal tolerance (UTT) CS, (C) UTT no tail flipping (NTF), and (D) UTT NPB in E. marinus. Data points represent individual measurements. The Pearson correlation coefficient and degrees of freedom, together with the probability values (P), are provided. For significant relationships, a line representing the linear relationship is shown.

9

Global change and the individual approach Table 2 Comparative summary of the outcome of the analyses of independent samples and analyses of individuals, for all traits investigated Type of analysis Naþ/Kþ-ATPase activity

Independent

Individual

Temperature  pCO2

Temperature  pCO2  MO2

UTL continuous swimming

Temperature  pCO2

No effect

UTL no tail flipping

No effect

Temperature  pCO2  MO2

UTL no pleopod beating

No effect

MO2

temperature and pCO2, always explaining a large amount of the observed variation. Individual variation can be biological meaningful

Overall, our study highlights the importance and value of employing an individual approach, as opposed to an independent-samples approach for the investigation of complex climatic changes on organismal functions, even if the interpretation of the biological significance of the results obtained may be more complex, that is, as was the case in our study three-way interactions with a continuous variable rarely are simple to explain. However, this complexity emerges from the great variation in responses we record in different individuals from the same population, variation that should not be ignored but taken into account (Bennett 1987). Finally, as the majority of investigations undertaken to date within the context of global climatic change have used analyses of independent samples, there is a serious risk that our interpretation of organisms’ responses to complex climatic change is (at least in part) inaccurate, or even incorrect, i.e., the ‘‘direction,’’ ‘intensity,’’ and/or level of predictability of organisms’ responses under examination may differ from those reported.

Acknowledgments We thank the numerous members of Team Evolution who helped construct and maintain the experimental system, as well as collecting and culturing E. marinus. We also thank R. Haslam, A. Torr, and R. Ticehurst for technical support. This work was undertaken while P.C. was in receipt of a Research Council UK Research Fellowship to investigate ocean acidification at Plymouth University.

Funding This project was supported by a NERC grant NE/ H017127/1 (to J.I.S. and P.C.) and is a contribution to the Task 1.4 ‘‘Identify the potential for organism resistance and adaptation to prolonged CO2 exposure’’ of the NERC Consortium Grant ‘‘Impacts of

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Treating individual differences in organismal functions as biologically meaningful (and not just ‘‘noise’’) yields a greater understanding of the selection processes occurring under complex climatic changes. As selection will likely play an important role in defining species’ responses to complex environmental changes (Pespeni et al. 2013), it is imperative we understand and incorporate individual variation into our studies and analyses. This fundamental aspect of biological systems has not received the attention it deserves, even though it can help predicting which phenotypes will be more likely to cope with future environmental challenges. For instance, an individual’s MO2 relates strongly and positively to UTT (as NTF and NPB), as one may have expected based on the oxygen limitation hypothesis (see Po¨rtner 2001; Verberk and Bilton 2011). Similarly, an individual’s MO2 relates strongly and positively to activity of Naþ/Kþ-ATPase. Based on these findings, we may predict that individuals with higher MO2 should be better able to cope with global warming (Sokolov et al. 2009), climatic anomalies (IPCC 2012), oceanic acidification (Caldeira and Wickett 2003), and possibly even with fluctuations in salinity levels that are expected to increase in intensity and frequency with increasing incidences of flash floods (IPCC 2012). In this sense, given that the direction of selection for MO2 phenotypes leading to an improvement of UTT and Naþ/Kþ-ATPase performances is the same, we do not expect evolutionary trade-offs between these two functions to emerge. On the other hand, individuals with higher MO2 levels would be expected to have greater energy costs for maintenance and repair, with likely a lower fraction of energy available for growth and reproduction. Thus, trade-offs may still emerge between physiological functions and life history, via an energy-reallocation route (Sibly and Calow 1986).

Conclusions

10

ocean acidification on key benthic ecosystems, communities, habitats, species and life cycles.’’

References

DeFur PL, Mangum CP, Reese JE. 1990. Respiratory responses of the blue crab Callinectes sapidus to long-term hypoxia. Biol Bull 178:46–54. Depledge MH. 1994. Genotypic toxicity: implications for individuals and populations. Environ Health Perspect 102(Suppl. 12):101–4. Dickson AG. 1990. Thermodynamics of the dissociation of boric acid in synthetic seawater from 273.15 to 318.15 K. Deep-Sea Res 37:755–66. Dickson AG, Millero FJ. 1987. A comparison of the equilibrium constants for the dissociation of carbonic acid in seawater media. Deep-Sea Res 34:1733–43. Foo SA, Dworjanyn SA, Poore AGB, Byrne M. 2012. Adaptive capacity of the habitat modifying sea urchin Centrostephanus rodgersii to ocean warming and ocean acidification: performance of early embryos. PLoS One 7:e42497. Hagerman L. 1986. Haemocyanin concentration in the shrimp Crangon crangon (L.) after exposure to moderate hypoxia. Comp Biochem Physiol A 85:721–4. Hagerman L, Uglow RF. 1985. Effects of hypoxia on the respiratory and circulatory regulation of Nephrops norvegicus. Mar Biol 87:273–8. Hagerman L, Sondergaard T, Weile K, Hosie D, Uglow RF. 1990. Aspects of blood physiology and ammonia excretion of Nephrops norvegicus under hypoxia. Comp Biochem Physiol 97:51–5. Hammer KM, Pedersen SA, Størseth TR. 2012. Elevated seawater levels of CO2 change the metabolic fingerprint of tissues and hemolymph from the green shore crab Carcinus maenas. Comp Biochem Physiol D 7:292–302. Hanski I, Singer MC. 2001. Extinction-colonization dynamics and host-plant choice in butterfly metapopulations. Am Nat 158:341–53. Hauton C, Tyrrell T, Williams J. 2009. The subtle effects of sea water acidification on the amphipod Gammarus locusta. Biogeosciences 6:1479–89. Hayes JP, Garland TJ, Dohm MR. 1992. Individual variation in metabolism and reproduction of Mus: are energetics and life history linked? Funct Ecol 6:5–14. IPCC. 2012. Managing the risks of extreme events and disasters to advance climate change adaptation. In: Field CB, Barros V, Stocker TF, Qin D, Dokken DJ, Ebi KL, Mastrandrea MD, Mach KJ, Plattner G-K, Allen SK, et al., editors. Managing the risks of extreme events and disasters to advance climate change adaptation. Cambridge, UK: Cambridge University Press. 528 p. Joyce AE. 2006. The coastal temperature network and ferry route programme: long-term temperature and salinity observations. Science Series Data Report. Lowestoft, UK: Cefas. Kelly MW, Hofmann GE. 2012. Adaptation and the physiology of ocean acidification. Funct Ecol published online (http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2435. 2012.02061.x/pdf). Lincoln RJ. 1979. British marine Amphipoda: Gammaridea. London, UK: British Museum (Natural History). p. 658. Lutterschmidt WI, Hutchison VH. 1997a. The critical thermal maximum: history and critique. Can J Zool 75:1561–74. Lutterschmidt WI, Hutchison VH. 1997b. The critical thermal maximum: data to support the onset of spasms as the definitive end point. Can J Zool 75:1553–60.

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Agnew DJ, Taylor AC. 1986. Seasonal and diel variations of some physiochemical parameters of boulder shore habitats. Ophelia 25:83–95. Aldrich JC. 1975. Individual variability in oxygen consumption rates of fed and starved Cancer pagurus and Maia squinado. Comp Biochem Physiol 51:175–83. Aldrich JC. 1989. Phenotypic responses and individuality in aquatic ectotherms. Wicklow, Ireland: JAPAGA. p. 212. Baden SP, Pihl L, Rosenberg R. 1990. Effects of oxygen depletion on the ecology, blood physiology and fishery of the Norway lobster Nephrops norvegicus. Mar Ecol Prog Ser 67:141–55. Bennett AF. 1987. Chapter 7—interindividual variability: an underutilized resource. In: Feder ME, Bennett AF, Burggren WW, Huey RB, editors. New directions in ecological physiology. New York (NY): Cambridge University Press. p. 147–69. Bolnick DI, Amarasekare P, Ara0jo MS, Bu¨rger R, Levine JM, Novak M, Rudolf VHW, Schreiber SJ, Urban MC, Vasseur DA. 2011. Why intraspecific trait variation matters in community ecology. Trends Ecol Evol 26:183–92. Bradford MM. 1976. Rapid and sensitive method for quantitation of microgram quantities of protein utilizing principle of protein-dye binding. Anal Biochem 72:248–54. Brooks SJ, Lloyd Mills C. 2003. The effect of copper on osmoregulation in the freshwater amphipod Gammarus pulex. Comp Biochem Physiol A 135:527–37. Brown JH, Gillooly JF, Allen AP, Savage VM, West GB. 2004. Toward a metabolic theory of ecology. Ecology 85:1771–89. Burton T, Killen SS, Armstrong JD, Metcalfe NB. 2011. What causes intraspecific variation in resting metabolic rate and what are its ecological consequences? Proc R Soc Lond B 278:3465–73. Caldeira K, Wickett M. 2003. Anthropogenic carbon and ocean pH: the coming centuries may see more ocean acidification than the past 300 million years. Nature 425:365. Calosi P, Bilton DT, Spicer JI, Atfield A. 2008. Thermal tolerance and geographical range size in the Agabus brunneus group of European diving beetles (Coleoptera : Dytiscidae). J Biogeogr 35:295–305. Calosi P, Ugolini A, Morritt D. 2005. Physiological responses to hyposmotic stress in the supralittoral amphipod Talitrus saltator (Crustacea: Amphipoda). Comp Biochem Physiol 142:267–75. Calow P, Forbes VE. 1998. How do physiological responses to stress translate into ecological and evolutionary processes. Comp Biochem Physiol A 120:11–16. Cianciaruso MV, Batalha MA, Gaston KJ, Petchey OL. 2009. Including intraspecific variability in functional diversity. Ecology 90:81–9. Crnokrak P, Roff DA. 2002. Trade-offs to flight capability in Gryllus firmus: the influence of whole-organism respiration rate on fitness. J Evol Biol 15:388–98. Darwin C. 1859. On the origin of species by mean of natural selection, or the preservation of favoured races in the struggle for life. London, UK: John Murray.

P. Calosi et al.

Global change and the individual approach

Rezende EL, Tejedo M, Santos M. 2011. Estimating the adaptive potential of critical thermal limits: methodological problems and evolutionary implications. Funct Ecol 25:111–21. Schlegel P, Havenhand JN, Gillings MR, Williamson JE. 2012. Individual variability in reproductive success determines winners and losers under ocean acidification: a case study with sea urchins. PLoS One 7:e53118. Senkbeil EG, Wriston JC. 1980. Hemocyanin synthesis in the American lobster, Homarus americanus. Comp Biochem Physiol B 68:163–71. Sibly RM, Calow P. 1986. Physiological ecology of animals: an evolutionary approach. Hong Kong, China: Blackwell Scientific Publications. p. 179. 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 crab Necora puber. Aquat Biol 10:11–21. Sokolov AP, Stone PH, Forest CE, Prinn R, Sarofim MC, Webster M, Paltsev S, Schlosser CA, Kicklighter D, Dutkiewicz S, et al. 2009. Probabilistic forecast for 21st century climate based on uncertainties in emissions (without policy) and climate parameters. Report No. 169. Cambridge (MA): MIT Joint Program on the Science and Policy of Global Change. Spicer JI, Baden SP. 2001. Environmental hypoxia and haemocyanin between-individual variability in Norway lobsters Nephrops norvegicus (L.). Mar Biol 139:727–34. Spicer JI, Gaston KJ. 1999. Physiological diversity and its ecological implications. Oxford, UK: Blackwell Science. p. 252. Sunday JM, Crim RN, Harley CDG, Hart MW. 2011. Quantifying rates of evolutionary adaptation in response to ocean acidification. PLoS One 6:e22881. Terblanche JS, Deere JA, Clusella-Trullas S, Janion C, Chown SL. 2007. Critical thermal limits depend on methodological context. Proc R Soc Lond B 274:2935–42. Thomsen J, Melzner F. 2010. Moderate seawater acidification does not elicit long-term metabolic depression in the blue mussel Mytilus edulis. Mar Biol 157:2667–76. Verberk WCEP, Bilton DT. 2011. Can oxygen set thermal limits in an insect and drive gigantism? PLoS One 6:e22610. Ve´zina F, Speakman JR, Williams TD. 2006. Individually variable energy management strategies in relation to energetic costs of egg production. Ecology 87:2447–58. Whiteley NM. 2011. Physiological and ecological responses of crustaceans to ocean acidification. Mar Ecol Prog Ser 430:257–71. Widdicombe S, Spicer JI. 2008. Predicting the impact of ocean acidification on benthic biodiversity: what can animal physiology tell us? J Exp Mar Biol Ecol 366:187–97. Widdicombe S, Dashfield SL, McNeil CL, Needham HR, Beesley A, McEvoy A, Øxnevad S, Clarke KR, Berge JA. 2009. Effects of CO2 induced seawater acidification on infaunal diversity and sediment nutrient fluxes. Mar Ecol Prog Ser 379:59–75.

Downloaded from http://icb.oxfordjournals.org/ by guest on May 10, 2013

Massamba-N’Siala G, Calosi P, Bilton DT, Prevedelli D, Simonini R. 2012. Life-history and thermal tolerance traits display different thermal plasticities and relationships with temperature in the marine polychaete Ophryotrocha labronica La Greca and Bacci (Dorvilleidae). J Exp Mar Biol Ecol 438:109–17. Mehrbach C, Culberson CH, Hawley JE, Pytkowicz RM. 1973. Measurement of the apparent dissociation constants of carbonic acid in seawater at atmospheric pressure. Limnol Oceanogr 18:897–907. Melatunan S, Calosi P, Rundle SD, Moody AJ, Widdicombe S. 2011. Exposure to elevated temperature and pCO2 reduces respiration rate and energy status in the periwinkle Littorina littorea. Physiol Biochem Zool 84:583–94. Melatunan S, Calosi P, Rundle SD, Widdicombe S, Moody JA. 2013. The effects of ocean acidification and elevated temperature on shell plasticity and its energetic basis in an intertidal gastropod. Mar Ecol Prog Ser 472:155–68. Melzner F, Go¨bel S, Langenbuch M, Gutowska MA, Po¨rtner H-O, Lucassen M. 2009. Swimming performance in Atlantic Cod (Gadus morhua) following long-term (4–12 months) acclimation to elevated seawater pCO2. Aquat Toxicol 92:30–7. Morris S, Taylor AC. 1983. Diurnal and seasonal variation in physico-chemical conditions within intertidal rock pools. Estuar Coast Shelf Sci 17:339–55. Newell RC. 1979. The biology of intertidal animals. Faversham, UK: Marine Ecological Surveys. p. 781. Pespeni MH, Sanford E, Gaylord B, Hill TM, Hosfelt JD, Jaris HK, LaVigne M, Lenz EA, Russell AD, Young MK, et al. 2013. Evolutionary change during experimental ocean acidification. Proc Natl Acad Sci USA 110:6937–42. Pierrot D, Lewis E, Wallace DWR. 2006. MS Excel program developed for CO2 system calculations, ORNL/CDIAC-105. Oak Ridge National Laboratory, U.S. Department of Energy, Oak Ridge (TN): Carbon Dioxide Information Analysis Center. Pistevos JCA, Calosi P, Widdicombe S, Bishop JDD. 2011. Will variation among genetic individuals influence species responses to global climate change? Oikos 120:675–89. Po¨rtner H-O. 2001. Climate change and temperature-dependent biogeography: oxygen limitation of thermal tolerance in animals. Naturwissenschaften 88:137–46. Po¨rtner H-O, Bennett AF, Bozinovic F, Clarke A, Lardies MA, Lucassen M, Pelster B, Schiemer F, Stillman JH. 2006. Trade-offs in thermal adaptation: the need for a molecular to ecological integration. Physiol Biochem Zool 79:295–313. Prosser CL. 1991. Environmental and metabolic animal physiology. New York (NY): Willey-Liss Inc. p. 592. Rastrick SPS, Whiteley NM. 2011. Congeneric amphipods show differing abilities to maintain metabolic rates with latitude. Physiol Biochem Zool 84:154–65. Re´ale D, Dingmanse NJ, Kazem AJN, Wright J. 2010. Evolutionary and ecological approaches to the study of personality. Phil Trans R Soc Lond B 365:3937–46.

11