INTRODUCTION
Studies of marine ecology and ecosystem functioning naturally depend on the simplification of complex, dynamic real-world systems. An ecological approach to define the true place of organisms within marine communities is, understandably, often constrained by a lack of baseline data, methodological limitations, or even because of a bias towards study of more ‘attractive’ species. Gaps in knowledge about the precise trophic positions of, or intra-guild relationships between, common taxa often cause notable biases in ecosystem models (Pauly et al., Reference Pauly, Graham, Libralato, Morissette and Deng Palomares2009; Bolnick et al., Reference Bolnick, Amarasekare, Araújo, Bürger, Levine, Novak, Rudolf, Schreiber, Urban and Vasseur2011), and such organisms are sometimes amalgamated or omitted altogether. Some taxa, especially those that are morphologically similar, may be placed together within model units that reduce the true complexity of their ecology, life history or physiology (Padilla & Allen, Reference Padilla and Allen2000). As these models often form the basis of predictions of the impacts of environmental change, undocumented and unconsidered variations in basic relationships and physiological traits may significantly affect such predictions (Boero et al., Reference Boero, Bouillon, Gravili, Miglietta, Parsons and Piraino2008; Pauly et al., Reference Pauly, Graham, Libralato, Morissette and Deng Palomares2009; Bolnick et al., Reference Bolnick, Amarasekare, Araújo, Bürger, Levine, Novak, Rudolf, Schreiber, Urban and Vasseur2011).
Polyplacophoran molluscs (or chitons) are one such group sometimes regarded as homogeneous in studies of marine benthic communities (e.g. Steneck & Watling, Reference Steneck and Watling1982; Ortiz & Wolff, Reference Ortiz and Wolff2002; Bishop, Reference Bishop2003; Hetherington & Reid, Reference Hetherington and Reid2003), or often omitted altogether (e.g. Poloczanska et al., Reference Poloczanska, Smith, Fauconnet, Healy, Tibbetts, Burrows and Richardson2011). Chitons are characterized by their eight articulating shells, or valves, and usually feed as generalist grazers on hard substrates in the intertidal and shallow subtidal, with many species also present in deep-sea habitats. They are often present in high densities with numerous species co-occurring, and can form a significant component of marine benthic communities (Horn, Reference Horn1982; Paine, Reference Paine1992; Littler et al., Reference Littler, Littler and Taylor1995; Barbosa et al., Reference Barbosa, Byrne and Kelaher2008), often controlling algal biomass and thus community structure (Dethier & Duggins, Reference Dethier and Duggins1984, Reference Dethier and Duggins1988). In some regions chiton species may be amongst the most abundant shore molluscs (Boyle, Reference Boyle1970). Numerous chiton species apparently co-occur within ranges and habitats, and phylogeographical studies have addressed the recent and continuing radiation of species in apparently widely-overlapping ecological niches (Kelly & Eernisse, Reference Kelly and Eernisse2008).
Organisms living in the intertidal are likely to experience natural extremes of oxygen availability, whether within hypoxic sediments, in adverse water chemistry conditions, or through factors such as predator harassment causing valve closure and subsequent hypoxia within shells (Larade & Storey, Reference Larade, Storey, Storey and Storey2002). On rocky shores, tidepools and other microhabitats often experience extremes of conditions, from severe hypoxia at night to ambient oxygen tensions higher than air-saturation in daylight (Truchot & Duhamel-Jouve, Reference Truchot and Duhamel-Jouve1980; Hagerman, Reference Hagerman1998). Under the predicted impacts of climate change it is likely that all marine communities will experience conditions of decreased dissolved oxygen in the future (Brewer & Peltzer, Reference Brewer and Peltzer2009; Keeling et al., Reference Keeling, Koertzinger and Gruber2010). Physiological traits that confer better survival in oxygen-depleted conditions may have considerable impact on the composition of marine communities in such conditions.
Physiological traits are important factors in determining an organism's success in a dynamic environment; the ability to metabolize oxygen may determine fitness, spatial distribution, and the capacity of a species to adapt to changing conditions. Physiological flexibility is necessary for organisms to survive in habitats that exhibit extremes of oxygen availability, especially those species that lack the locomotory ability to move to another area (e.g. Hagerman, Reference Hagerman1998). Such flexibility in physiology has been demonstrated in many organisms, including bivalves (de Zwaan et al., Reference de Zwaan, Cortesi, Thillart, Roos and Storey1991; Shumway et al., Reference Shumway, Scott and Shick1993), gastropods (Eberlee & Storey, Reference Eberlee and Storey1988; Pannunzio & Storey, Reference Pannunzio and Storey1998), and crustacean zooplankton (McAllen et al., Reference McAllen, Taylor and Davenport1999).
Responses to hypoxic conditions differ between invertebrates. Motile species such as littorinid snails may simply relocate to escape hypoxia, while more sedentary species such as bivalves instead rely on a reduction in metabolic rate (Livingstone, Reference Livingstone and Gilles1991). One possible physiological response to hypoxia is a compensatory increase in the uptake rate of oxygen in order to maintain adequate supply (Mangum & Van Winkle, Reference Mangum and Van Winkle1973). Many species lack the physiological mechanisms necessary to maintain oxygen uptake under hypoxic conditions and as a result, oxygen uptake is proportional to ambient oxygen tension (oxyconformers) (e.g. Wilson & Davis, Reference Wilson and Davis1984; Spicer et al., Reference Spicer, Dando and Maltby2002). However some species, often those found where large variations in oxygen tensions are common, can maintain routine oxygen uptake rates despite declining oxygen tension (oxyregulators) (e.g. Taylor & Moore, Reference Taylor and Moore1995; Strahl et al., Reference Strahl, Dringen, Schmidt, Hardenberg and Abele2011). Typically this occurs over a range of oxygen tensions to a critical tension (Pcrit) below which they display oxyconforming behaviour. The mechanisms by which this increase may occur are varied and may include increases in ventilation of the gills (Taylor & Moore, Reference Taylor and Moore1995), or in species with active circulatory systems an increase in heart rate or heart stroke volume (Bayne, Reference Bayne1971; DeFur & Mangum, Reference De Fur and Mangum1979). Gastropods such as the limpets Patella granularis Linnaeus, 1758 and Siphonaria capensis Quoy & Gaimard, 1833 may induce bradycardia as a response in order to decrease short-term oxygen consumption (Marshall & McQuaid, Reference Marshall and McQuaid1993).
In chitons, variation in oxygen consumption has been observed within a single species (Chiton pelliserpentis Quoy & Gaimard, 1835) related to size, temperature and shore position (Horn, Reference Horn1985). Another species, Chiton stokesii Broderip & Sowerby, 1832 appears to possess physiological adaptations to extended periods of emersion to better exploit the intertidal habitat (McMahon et al., Reference McMahon, Burggren, Pinder and Wheatly1991). Only one previous study has examined the variation in oxygen metabolism between co-occurring species of chiton (Murdoch & Shumway, Reference Murdoch and Shumway1980). Chitons make a good candidate to examine how a single physiological trait may vary between species that are often assumed to have identical roles in ecosystem functioning, and so demonstrate how intra-guild complexity is often neglected at larger scales. Here we present a case study to examine the variation in oxygen uptake within a morphologically constrained group of common, co-occurring intertidal molluscs, within and between Atlantic and Pacific temperate ecosystems.
MATERIALS AND METHODS
Two sets of experiments were conducted to examine standard respiration rates and responses to declining oxygen tensions in Atlantic (Ireland) and Pacific (British Columbia, Canada) species of chiton.
Atlantic experiments
Three species of chiton, Lepidochitona cinerea Linnaeus, 1767 (N = 10), Acanthochitona crinita Pennant, 1777 (N = 15) and Leptochiton asellus Gmelin, 1791 (N = 5) (Figure 1), were collected from the low intertidal zone at six sites in Strangford Lough, Northern Ireland, October 2010. Some Leptochiton asellus specimens were collected by SCUBA divers from subtidal Modiolus modiolus Linnaeus, 1758 beds at 10–20 m depth. Specimens were housed in aerated seawater obtained from Strangford Lough (14°C, salinity = 33.5), and kept attached to bare rocks but otherwise not actively fed. Experiments took place from October to December 2010.
Chitons examined for oxygen uptake rates in this study. Atlantic species: (A) Acanthochitona crinita; (B) Lepidochitona cinerea; (C) Leptochiton asellus. Pacific species: (D) Leptochiton rugatus; (E) Tonicella lineata; (F) Mopalia ferreirai. All animals are shown with anterior to the right; all scale bars are 10 mm.
On the day of each experiment specimens were transferred to a holding vessel containing 22 µm-filtered seawater at 14°C and cleaned gently with a fine, soft brush to remove epibiota. They were then individually transferred to glass respirometers fitted with a Clark oxygen electrode and a water jacket kept at a temperature of 14°C. A stir bar in the base of the chamber was used to ensure adequate mixing in the respirometer, and was guarded by a plastic barrier to prevent interference by the specimen. The chamber was sealed with a Perspex stopper possessing a narrow (<1 mm) central aperture, allowing the pressure within the experimental chamber to remain equalized with the external atmospheric pressure. Changes in oxygen tension were monitored by a PC recording module at a sampling interval of 100 ms. Trials were continued until oxygen was exhausted or uptake had noticeably levelled off.
After removal from the chamber, specimens were blotted dry on tissue paper, and wet mass determined. Before each trial the Clark electrode was calibrated for zero and saturated oxygen concentrations using anoxic and fully air-saturated seawater. After experiments the chamber was cleaned with filtered seawater and the calibrations checked, with any drift of sensitivity noted, and the apparatus recalibrated if necessary for the next experimental trial. In the interests of data integrity, experimental runs with a drift of more than 10% in sensitivity were discarded. Drift in probe sensitivity below this threshold was assumed to be linear over the course of the experiment, and the probe recordings adjusted as such.
Specimens were dried at 60°C until constant mass was achieved to obtain total dry mass (TDM). They were then incinerated at 500°C for two hours in a muffle furnace to obtain shell and ash mass, and this subtracted from TDM to obtain the ash-free dry mass of tissue (AFDM). All masses are reported in grams.
Pacific experiments
All specimens were collected intertidally in British Columbia, Canada, June 2008; Leptochiton rugatus Carpenter in Pilsbry, 1892 (N = 10), at Whiffen Spit, Sooke, Vancouver Island; Tonicella lineata Wood, 1815 (N = 8) and Mopalia ferreirai Clark, 1991 (N = 3) at Walker's Hook, Salt Spring Island (Figure 1). Specimens were transported to University of British Columbia (Vancouver) aquarium facilities and maintained in aerated, filtered seawater (12°C, salinity = 32) for eight days before experiments and not fed during this time. On the day of each trial four specimens were moved to a water-bath with the same salinity and temperature as the acclimation aquarium. Chitons were individually placed in glass respirometers of a volume matching roughly 20 times that of the chiton, with a magnetic stir bar. The respirometry chambers were sealed with a rubber stopper, fixed with parafilm, and fitted with a fibre-optic oxygen probe (FOXY systems, Ocean Optics, Dunedin, Florida) connected to a PC recording module. Four trials were run simultaneously and oxygen readings recorded at intervals of 15 seconds. Each trial was run for a minimum of three hours and at the end of each trial the wet mass of the chiton and the volume of the respirometer recorded, with additional trials that demonstrated leakage discarded.
Data analysis
Data were analysed using Microcal Origin 8.0 and probe recordings smoothed using Savitzky–Golay smoothing. Statistical tests were performed using R: a language and environment for statistical computing (R Core Development Team, 2012). Calibrations for zero and air-saturated oxygen tensions were used to convert oxygen probe recordings to oxygen concentrations using calculated air-saturated concentrations of 8.38 mg l−1 for the Atlantic seawater and 8.88 mg l−1 for the Pacific seawater. Air-saturated concentrations were determined according to the methods of Benson & Krause (Reference Benson and Krause1984).
Initial mass-specific basal oxygen uptake rates (VO2) (μgO2 min−1 g−1) were calculated for each specimen, based on the mean rate at which oxygen was depleted from fully air-saturated to 90% air-saturated, per time (minutes) and mass (AFDM) (Table 1). For Pacific specimens, only wet tissue mass was determined (by subtracting dry shell mass from total wet mass), and basal mass-specific uptake rates based on these masses will be used for analysis.
Mean values (±SE) for total wet mass (WM), wet tissue mass (WT), and basal oxygen uptake rates (VO2) for each species, standardized by each wet body mass metric. These are presented for comparisons of Pacific species examined here with historical data (Figure 5). VO2 for Atlantic species standardized by ash-free dry tissue mass are shown in Figure 4.
For further analysis in conditions of progressive hypoxia, uptake rates were standardized to a percentage, with basal VO2 of each specimen assigned a value of 100% and a zero uptake value assigned a value of 0% (standardized VO2 or SVO2). Rates of half (SVO250%) and one-quarter (SVO225%) these standardized basal rates were determined by plotting the rate of oxygen uptake against declining oxygen concentrations for each specimen and determining the concentrations at which SVO250% and SVO225% occurred. Where SVO250% and SVO225% occurred over a range of concentrations a mean of these was taken.
To quantify the degree of oxygen dependence of the different species we use the method described by Alexander & McMahon (Reference Alexander and McMahon2004) to determine the ‘regulation value’, R. Here, the integrated sum of each proportional value of VO2 at each 5% decrease in oxygen concentration is expressed as a percentage of that which would be expected for an organism that exhibits perfect oxygen uptake regulation. An idealized oxyregulator would have an R value of 100%, while an idealized oxyconformer (as represented by the solid lines in Figure 2) would have an R value of 50%. Values of R within this range represent different degrees of oxyregulation ability in a particular specimen; the closer these R values are to 100% the greater this ability (Alexander & McMahon, Reference Alexander and McMahon2004; Lencioni et al., Reference Lencioni, Bernabò, Vanin, Di Muro and Beltramini2008). In addition we calculated mean R 25% for each species, which is the respiration rate at oxygen concentrations 25% that of air-saturated concentrations, expressed as a percentage of the SVO2. Typical values for defining oxyregulating organisms are R > 50%, and R 25% > 37% (Alexander & McMahon, Reference Alexander and McMahon2004; Brodersen et al., Reference Brodersen, Pedersen, Lindegaard and Hamburger2004; Lencioni et al., Reference Lencioni, Bernabò, Vanin, Di Muro and Beltramini2008).
Standardized basal oxygen uptake rates for Atlantic (Figure 2A) and Pacific (Figure 2B) species in conditions of decreasing oxygen tensions. Vertical axis indicates basal rate of uptake for each specimen standardized to 100% (SVO2), with subsequent uptake rates calculated as a percentage of this at each 5% decrease in oxygen concentration (±SE). The metrics SVO250% and SVO225% are used in the text to indicate rates of half- and one-quarter basal mass-specific uptake rates respectively. Solid diagonal lines represent idealized oxyconforming behaviour in which oxygen uptake would be directly proportional to concentration. Deviation above this line indicates different degrees of oxyregulation ability (deviation below the line would indicate an organism with VO2 highly constrained by hypoxia). An idealized oxyregulator would be represented by a horizontal line at SVO2 across all concentrations, indicating an ability to maintain the same rate of uptake at any concentration of oxygen.
We also compare our results with those in published literature for chiton oxygen uptake rates. When necessary, data were extracted from appropriate figures using PlotDigitizer 2.5.1 for Mac OS X, and O2 concentration and uptake rate units converted to μgO2 min−1 g−1 according to the methods of García & Gordon (Reference García and Gordon1992). Mass-specific VO2 recorded in this study were recalculated and standardized by total wet mass, total wet tissue mass, or total dry tissue mass for comparisons with historical data where appropriate. In chitons, ratios of the contribution of shell, tissue and water content to total mass vary considerably between species (McMahon et al., Reference McMahon, Burggren, Pinder and Wheatly1991). Previous studies have used a variety of mass measurements for calculating VO2, and we are careful to only make comparisons with historical data where mass measurements are comparable.
All tests of significance are one-way analyses of variance (ANOVAs) (α = 0.05), and all measurements of variability are standard error (SE). Post-hoc tests are Tukey multiple comparisons with 95% confidence intervals. Data were tested to ensure they meet the assumptions of the ANOVA test (i.e. independence, normality and homogeneity of variances).
RESULTS
The chiton species examined showed substantial variation in oxygen consumption rate, both in basal uptake under initial normoxic conditions, and in response to decreasing oxygen availability.
Changes in VO2 in response to decreasing oxygen concentrations were different amongst both the Atlantic and Pacific faunas (Figure 2). Among the Atlantic species, A. crinita was the strongest oxyregulator, with L. cinerea and Leptochiton asellus having mean regulation values of R and R 25% much closer to typical oxyconformer values (Table 2). Among the Pacific species, Leptochiton rugatus can be considered to be a typical oxyconformer, while M. ferreirai and T. lineata both showed strong oxyregulatory ability (Table 2).
Quantification of oxyregulatory ability for each species; mean R values and mean R 25% values (±SE). R values were determined according the method described by Alexander & McMahon (Reference Alexander and McMahon2004). R 25% indicates the respiration rate at 25% of air-saturated oxygen concentration as a percentage of SVO2. Typical values for defining oxyregulating organisms are R > 50%, and R 25% > 37% (Alexander & McMahon, Reference Alexander and McMahon2004; Brodersen et al., Reference Brodersen, Pedersen, Lindegaard and Hamburger2004; Lencioni et al., Reference Lencioni, Bernabò, Vanin, Di Muro and Beltramini2008).
In order to demonstrate the degree of variation between species in basic respiratory patterns independent of body mass, the mean oxygen concentrations at which SVO2 had declined to half (SVO250%) and one-quarter (SVO225%) of basal initial rates were determined (Figure 3). These metrics are mass-independent and so can be used to compare respiration patterns between individuals of different sizes (Spearman rank correlations: SVO250% to gWT rs = –0.54, N = 51, P < 0.0001; SVO225% to gWT rs = –0.52, N = 49, P = 0.0001). These metrics again demonstrated differences in oxyregulatory behaviour with decreasing oxygen concentration. SVO250% differed significantly between the six study species (F5,45 = 12.21, P < 0.0001: Figure 3A). The three species with high R values, A. crinita, M. ferreirai and T. lineata, showed different behaviour than the other species, as indicated by the lower mean concentrations at which SVO250% occurred, again indicating these three species were able to maintain higher relative oxygen uptake rates at lower concentrations than the other species, and so demonstrated stronger oxyregulatory ability.
Mean concentrations of O2 (mg l−1) at which SVO250% (Figure 3A) and SVO225% (Figure 3B) occur for Atlantic (white columns) and Pacific (grey columns) species examined in this study. The metrics SVO250% and SVO225% indicate rates of 50% and 25% of initial standardized basal rates respectively. Error bars indicate standard error.
Analysis within the respective geographical species groups also showed significant differences in the SVO250% metric (Atlantic species F2,27 = 12.59, P = 0.0001; Pacific species F2,18 = 12.91, P = 0.0003). Post-hoc tests reveal support for the categorization of the different species into groups with differing degrees of oxyregulatory ability. Tests show the single species with the strongest oxyconforming behaviour, Leptochiton rugatus to be significantly different from all other species (P < 0.0001). There were no significant differences between the species that apparently demonstrated oxyregulation: A. crinita, T. lineata and M. ferreirai (P = 0.45).
The SVO225% metric also differed significantly between the study species (F5,43 = 6.79, P = 0.0001: Figure 3B), though this was clearly a result of the relatively higher mean concentration at which SVO225% occurs for Leptochiton rugatus. Post-hoc Tukey tests revealed significant differences only involving Leptochiton rugatus, and with the exception of this species there was not a significant difference in SVO225% between the remaining species.
Basal mass-specific oxygen uptake rates under normoxic conditions (VO2) were significantly different between all six species examined using wet tissue mass (WT) as the standardizing mass denominator (F5,45 = 14.57, P < 0.0001). Lepidochitona cinerea in particular showed a markedly greater mean VO2 than any other species, the next closest being A. crinita (Table 1). Post-hoc Tukey tests indicate differences were due to the greater VO2 demonstrated by L. cinerea.
Within the Atlantic fauna there were significant differences when both wet tissue mass (WT) and ash-free dry tissue mass (AFDM) were used as the standardizing mass denominator; (F2,27 = 13.74, P = 0.0001) and (F2,27 = 16.19, P < 0.0001) respectively. Again, post-hoc Tukey tests reveal differences were driven by the greater VO2 demonstrated by L. cinerea, with no significant differences between the other two species.
The basal VO2 standardized by dry tissue mass of the Atlantic species examined in this study were compared to that of six New Zealand species (data extracted from Murdoch & Shumway (Reference Murdoch and Shumway1980, Figure 3) and Horn (Reference Horn1985, figures 1&3)). The New Zealand species demonstrate a similar level of variation in VO2 between species as was observed in this study (Figure 4). There was a significant difference in the VO2 within the New Zealand fauna (F6,124 = 35.98, P < 0.0001) and between all nine species (F9,151 = 31.52, P < 0.0001). Post-hoc Tukey tests show significant differences to be associated with L. cinerea and O. neglectus, which are different from all other species except for each other and other significant differences are associated with A. crinita and C. pelliserpentis (Figure 4).
Mean basal mass-specific VO2 (μgO2 min−1 gDM−1) for three Atlantic species in this study (white columns) and six New Zealand species examined by Murdoch & Shumway (Reference Murdoch and Shumway1980, figure 3), one of which was also examined by Horn (Reference Horn1985, figures 1&3) (grey columns). For Atlantic species ash-free dry tissue mass (AFDM) was used as the mass-standardizing denominator (DM). For DM, Murdoch & Shumway (Reference Murdoch and Shumway1980) used dry tissue mass as determined through subtraction of KOH-dissolved mass from total dry mass; Horn (Reference Horn1985) does not specify a method for determining DM. This study used the period over which oxygen concentration was reduced to 90% of air-saturated to determine basal VO2. Horn (Reference Horn1985) determined VO2 by averaging rates over a 2-hour period. Murdoch & Shumway (Reference Murdoch and Shumway1980) do not specify a methodology or time period over which uptake rates were determined. Temperatures at which species were examined are: this study 14°C; Murdoch & Shumway (Reference Murdoch and Shumway1980) 15°C; Horn (Reference Horn1985) 15.5°C. Columns are arranged with low shore or subtidal species on the left, with those to the right occurring progressively higher in the intertidal (within the respective geographical groups). Error bars indicate standard error.
There was no significant difference in basal VO2 between the three Pacific species examined in this study (Table 1). Comparisons with historical data from other Pacific species are presented in Figure 5, using data extracted from appropriate tables, figures or text. However none of these studies provide data with which to further test statistical differences in VO2 between Pacific species.
Mean basal mass-specific VO2 standardized by total wet mass (μgO2 min−1 gWM−1) for three Pacific species in this study (white columns), and additional Pacific species taken from existing literature (grey columns). K, Kincannon (Reference Kincannon1975, figures 3&4: T, 13°C); R, Robbins (Reference Robbins1975, table 1: T, 13.5°C); L, Lebsack (Reference Lebsack1975, figure 1: T, 13.5°C); S&S, Stickle & Sabourin (Reference Stickle and Sabourin1979, p. 266: T, 13°C); R&S, Rostal & Simpson (Reference Rostal and Simpson1988, p.124: T, 11°C). Species in this study were examined at 12°C. Error bars indicate standard error and are presented only when they could be extracted from the historical data.
DISCUSSION
Simplification and abstraction of complex and dynamic environmental systems is necessary to make sense of marine ecology. However, such simplification must strike a balance between practicality and true reflection of real-world complexity (Warwick, Reference Warwick1993), and seek to fully incorporate previously neglected aspects of the ecology and biology of component taxa (Boero et al., Reference Boero, Bouillon, Gravili, Miglietta, Parsons and Piraino2008). Given the morphological and ecological similarity of chitons, there is a surprisingly large variation in uptake rates and behaviour among species globally, and co-occurring species in separate faunas demonstrate substantial differences in physiology.
Comparative data
A small number of other studies have quantified respiratory physiology in chitons. Each of these employed different experimental and analytical methods, were conducted at different times in the seasonal cycle or at different temperatures, and mostly were restricted to manipulations of one or two taxa. Variations in metabolic rates due to extrinsic factors such as time of year and temperature are to be expected, so direct comparisons of uptake rates should be considered with this caveat in mind. The two groups of species examined in the present study were examined at different times of year and under slightly different laboratory conditions, which may account for some differences in respiratory rates. However, basic physiological traits such as oxyregulatory ability are likely to be inherent (Mangum & Van Winkle, Reference Mangum and Van Winkle1973). Nagabhushanam & Murti (Reference Nagabhushanam and Murti1972) reported effects of body size and salinity on respiration in Chiton granoradiatus Leloup, 1937 in India. Petersen & Johansen (Reference Petersen and Johansen1973) measured metabolism in a range of body sizes and temperatures of adult Cryptochiton stelleri von Middendorff, 1847 in the north-east Pacific. Kincannon (Reference Kincannon1975) reported oxygen consumption in T. lineata (included here as one of our Pacific species) from intertidal and subtidal populations, and at different temperatures. Robbins (Reference Robbins1975) also examined aerial and aquatic respiration in T. lineata, as well as Nuttallina californica Nuttall MS, Reeve, 1847. Lebsack (Reference Lebsack1975) and Stickle & Sabourin (Reference Stickle and Sabourin1979) reported data for respiration varying with temperature and salinity in two common species in the north-east Pacific (Mopalia muscosa Gould, 1846 and Katharina tunicata Wood, 1815, respectively). Rostal & Simpson (Reference Rostal and Simpson1988) also examined K. tunicata in different salinities. Horn (Reference Horn1985) recorded temperature sensitivity in the respiratory rates of Chiton pelliserpentis at varying shore heights, but no overall difference in rates. This was also one of the species examined by Murdoch & Shumway (Reference Murdoch and Shumway1980), who tested six co-occurring species of chitons in New Zealand from different shore heights, in terms of their oxygen uptake and oxyregulatory function. McMahon et al. (Reference McMahon, Burggren, Pinder and Wheatly1991) measured oxygen uptake in the tropical species Chiton stokesii in Panama. Not all of these publications include data that could be directly compared to our present results, but quantitative comparisons were made wherever appropriate conversions were determinable (Figures 4 & 5). Studies that used tropical study organisms (Nagabhushanam & Murti, Reference Nagabhushanam and Murti1972; McMahon et al., Reference McMahon, Burggren, Pinder and Wheatly1991) show much higher metabolism, as would be expected of ectothermic organisms in higher ambient temperatures. The more meaningful comparisons with our results are the global trends in animals at temperate latitudes.
Various approaches have been proposed to quantify the degree of oxyregulation an organism exhibits over a range of oxygen tensions; the ongoing methodological development being one aspect that confounds direct comparison with historical data. Categorization of organisms as either oxyconformers or oxyregulators is highly simplistic, as there is a continuum of oxyregulatory responses by species (Taylor & Brand, Reference Taylor and Brand1975; Alexander & McMahon, Reference Alexander and McMahon2004). Tang (Reference Tang1933) and Bayne (Reference Bayne1973) show how the intercept (K1) and slope (K2) of a linear regression of VO2 against oxygen concentration may be used to obtain an oxygen independence index (K1/K2) for comparison of regulatory ability between species. This method was used by Murdoch & Shumway (Reference Murdoch and Shumway1980) to demonstrate that the degree of oxyregulation in the chitons they examined appears to be greater the higher the species is found in the intertidal. Other methods use semi-logarithmic, exponential, or hyperbolic regressions (Mangum & Van Winkle, Reference Mangum and Van Winkle1973; Taylor & Brand, Reference Taylor and Brand1975; Herreid, Reference Herreid1980) to obtain coefficients with which to determine the degree of oxygen regulation. However, such calculations are often based on relatively few data points, and as such are not particularly suited to dense, high-resolution data (such as recorded in this study) where changes in uptake rates may show a strong trend but fluctuate significantly over short time periods. As Alexander & McMahon (Reference Alexander and McMahon2004) point out, some species are particularly adept at oxyregulation and in such cases regressions on which to base these metrics are difficult to fit. In some species uptake may vary unpredictably with progressive hypoxia, and may actually increase initially (Alexander & McMahon, Reference Alexander and McMahon2004). This was observed here in some A. crinita and T. lineata specimens (both oxyregulators) which showed initial increases in uptake rates as oxygen concentrations declined. While oxyregulatory ability is likely to be inherent in species (Mangum & Van Winkle, Reference Mangum and Van Winkle1973), the ability may not necessarily be observed in particular individuals for a number of reasons, and several studies have demonstrated that some individuals show strong oxyregulation while others of the same species act as conformers under similar conditions (Bayne, Reference Bayne1971; Herreid, Reference Herreid1980; Duke & Ultsch, Reference Duke and Ultsch1990). Some A. crinita specimens in this study (both small and large individuals within the sample set) showed strong oxyregulation while others did not (R values ranged from 87% to 61%). The strongly oxyregulating A. crinita specimens were able to maintain basal VO2 in ambient oxygen concentrations ranging from fully air-saturated (8.38 mg l1-) to approximately 35% of air-saturated (3.0 mg l−1). Anecdotally, we have also observed A. crinita specimens that were revived and apparently healthy after prolonged periods (over 36 hours) of exposure to almost totally anoxic conditions (Carey & Sigwart, unpublished observation). Other individual specimens of the same taxon showed generally linear, oxyconforming relationships to ambient oxygen tensions, similar to those observed in L. cinerea and Leptochiton asellus. No L. cinerea or Leptochiton asellus specimens showed evidence of substantial oxyregulation ability.
McMahon et al. (Reference McMahon, Burggren, Pinder and Wheatly1991) showed increases in haemoconcentration and haemolymph pressure of C. stokesii associated with air exposure, and similar mechanisms may be utilized by A. crinita and the other oxyregulating species to increase oxygen uptake in depleted conditions. Our data measured at the organismal level however, cannot indicate how or at what stage in the metabolic pathway oxyregulation is implemented.
Few other studies have quantified VO2 in chitons in a manner that can be directly compared to the species included here. Kincannon (Reference Kincannon1975, figures 3 & 4) quantified mean mass-specific VO2 in T. lineata (T = 13°C), with a result comparable with that found in this study given that our experiments were conducted at a slightly lower temperature (Figure 5). Robbins (Reference Robbins1975, table 1) however, found a higher mass-specific VO2 in T. lineata (T = 13.5°C) (Figure 5). The disparity in this result with that presented here and with Kincannon (Reference Kincannon1975) is too great to be explained solely by the slightly higher temperature. Robbins (Reference Robbins1975) found another co-occurring species N. californica to have a greater VO2 than recorded in any other temperate Pacific species. Lebsack (Reference Lebsack1975) and Stickle & Sabourin (Reference Stickle and Sabourin1979) found VO2 in two other Pacific species more comparable to those recorded in this study (Figure 5). Within the New Zealand fauna, Murdoch & Shumway (Reference Murdoch and Shumway1980) and Horn (Reference Horn1985) independently measured VO2 in C. pelliserpentis with results not significantly different (Figure 4). While there was no significant difference in VO2 between the Pacific species in this study, the historical data (Figure 5) suggests some variation may exist.
These existing studies illustrate the issues with comparisons of historical data. With the small number of datasets available it is not always clear whether there are unknown variables that complicate the results, or if one study is anomalous. In the case of chitons, several species complexes, including some of those examined in the present study such as T. lineata and several Mopalia species, have been redescribed since historical experiments were published (Eernisse et al., Reference Eernisse, Clark, Draeger and Carlton2007). There is no way of checking the modern identity of the species used in these historical studies.
Below we consider several specific potential factors (body size, shore height, local environmental conditions and phylogeny) that may control the variation observed among our study species and the few previously published observations.
Body size
The influence of body mass on oxygen consumption has been well studied in many marine species (Newell & Roy, Reference Newell and Roy1973; Fisher, Reference Fisher1976; Bridges & Brand, Reference Bridges and Brand1980; Katsanevakis et al., Reference Katsanevakis, Xanthopoulos, Protopapas and Verriopoulos2007; Seibel, Reference Seibel2007). The greater the body mass of an organism, the greater its oxygen consumption with this relationship following generalized mass-dependent scaling factors (West et al., Reference West, Brown and Enquist1997; Gillooly et al., Reference Gillooly, Brown, West, Savage and Charnov2001). In our study, the species that showed oxyregulatory ability (A. crinita (Atlantic), and M. ferreirai, and T. lineata (Pacific)) are also those with generally greater adult body mass of the sampled fauna, both observationally and among the individual specimens included in our experiments (Table 1). The species we examined that are typically smaller in body mass showed general oxyconforming behaviour, suggesting direct diffusion processes may dominate uptake, rather than active responses such as increased ventilation of gills or haemolymph circulation or pressure modifications. Those species with larger body mass may not be able to rely on diffusion alone to supply body tissues with oxygen, and so may possess additional mechanisms to regulate respiration. However, Cryptoconchus porosus Blainville MS, Burrow, 1815 in New Zealand, the largest species examined by Murdoch & Shumway (Reference Murdoch and Shumway1980) was also one of the strongest oxyconformers in their study.
Body mass is clearly a factor in the observed differences in total oxygen consumption between the species studied here. However, in our chosen metrics (SVO250% and SVO225%) there was no significant correlation with body mass, demonstrating that interspecific differences in oxygen physiology are a much greater factor than differences between individuals or species of different mass.
Shore height
Murdoch & Shumway (Reference Murdoch and Shumway1980, p. 132) state that oxygen demand is inversely related to shore height, and the species occurring highest on the shore have a lower rate of aquatic oxygen consumption than those lower down. While this is broadly true in their results, it takes no account of mass-specific differences in consumption. All of the species Murdoch & Shumway (Reference Murdoch and Shumway1980) examined were in approximately the same size-range, with the exception of C. porosus, which is generally larger, and had little overlap in mass with the other species they examined. The mass-specific uptake of C. porosus is actually among the lowest of those they examined (Figure 4), yet it occurs lower on the shore than the other species.
In order to relate shore position (or other abiotic factors) to oxygen consumption one must take account of relative body mass, which is an important confounding variable. The hypothesis put forward by Murdoch & Shumway (Reference Murdoch and Shumway1980) is not supported by the results described here. The highest occurring chiton species in the Atlantic fauna L. cinerea, had by far the greatest VO2 among those examined, and Leptochiton asellus, which is almost exclusively subtidal, had the lowest (Figure 4).
Similar to this study, Murdoch & Shumway (Reference Murdoch and Shumway1980) found one species, Onithochiton neglectus Rochebrune, 1881 to have a greatly higher mass-specific VO2 than the other species in the local fauna, and within the same general range of values that we observed in L. cinerea (Figure 4). In contrast however, O. neglectus is a low shore species, while L. cinerea is the highest occurring chiton species in the Atlantic fauna. Kincannon (Reference Kincannon1975) reported no significant difference in respiratory rates between sub- and intertidal T. lineata. Similarly Horn (Reference Horn1985) reported no difference in aquatic respiration rates between low and high shore C. pelliserpentis. Shore height is only one factor among many that may explain chiton respiratory behaviour and ability.
Murdoch & Shumway (Reference Murdoch and Shumway1980) also observed variations in oxyregulation between their species correlated with shore position, with species occurring higher up the shore possessing greater oxyregulatory ability. This agrees with findings in work on other taxa, which suggests a correlation between species possessing oxyregulatory ability and the possibility of experiencing hypoxia, such as those occurring higher in the intertidal (Sassaman & Mangum, Reference Sassaman and Mangum1972; Bayne, Reference Bayne1973). These results however, were not replicated here. Within the Atlantic species, A. crinita showed the greatest oxyregulatory ability, but it occurs lower on the shore than L. cinerea, which showed typical oxyconforming behaviour. Within our Pacific species there is broader agreement with the previous studies, with the two intertidal species possessing oxyregulatory ability, and the subtidal species being a typical oxyconformer.
Local environment
There was no significant difference in mass-specific VO2 among our Pacific species (Table 1), despite differences in shore height and also body mass. The total fauna of chiton species in the north-east Pacific is much more speciose than in the Atlantic and they also exhibit a much greater disparity in body mass (Kaas & Van Belle, Reference Kaas and Van Belle1985; Slieker, Reference Slieker2000; Eernisse et al., Reference Eernisse, Clark, Draeger and Carlton2007). A lack of variation in mass-specific VO2 might be indicative of an increase in niche specialization through other adaptations. Body mass modification may prove a more beneficial adaptation in exploiting new niche opportunities among a more competitive community than changes in basal physiological traits.
In the north-east Atlantic fauna, where there are fewer species, L. cinerea is by far the most common and widespread (Poppe & Goto, Reference Poppe and Goto1991), yet its mass-specific VO2 far exceeds that demonstrated by the other two Atlantic species in this study (Figure 4). This higher consumption may be indicative of an adaptation to a more active lifestyle unconstrained by competitors that has allowed it to exploit niche opportunities unavailable to the other species.
The two most common chitons on north-east Atlantic shores are A. crinita and L. cinerea, and while there are areas where one or other species may be more abundant, they are frequently found together in mixed communities. The ability of A. crinita to oxyregulate with increasing hypoxia is more typical of the physiological capacity required to survive the stresses of the intertidal zone, yet L. cinerea is much more widely distributed on Atlantic coasts and is found higher in the intertidal (Nichols, Reference Nichols1900; Poppe & Goto, Reference Poppe and Goto1991; Hayward & Ryland, Reference Hayward and Ryland1996). The greatly higher basal VO2 observed in L. cinerea might be indicative of a higher general metabolism and more active lifestyle. The microhabitat where L. cinerea is characteristically found is within cobble, which lacks the wide variety and abundance of biota found on the larger stationary boulders A. crinita frequents (Carey & Sigwart, unpublished observations). Many molluscs, including chitons, have a home ‘range’ over which they forage before returning to a home scar (e.g. Chelazzi et al., Reference Chelazzi, Focardi, Deneubourg, Chelazzi and Vannini1988). We speculate that the higher metabolism of L. cinerea may be related to the need to cover a greater range of this less productive habitat in order to maintain adequate nutrition. Wide variations in metabolic rate related to locomotory capacity have previously been observed within molluscs, albeit in the highly morphologically varied Cephalopoda (Seibel, Reference Seibel2007).
The seeming inability of L. cinerea to oxyregulate as effectively as other species might be related to this relatively higher VO2. Lepidochitona cinerea has a higher mass-specific VO2 of any species examined here or in any previous studies of temperate-latitude chitons (Kincannon, Reference Kincannon1975; Robbins, Reference Robbins1975; Murdoch & Shumway, Reference Murdoch and Shumway1980; Horn, Reference Horn1985), and has been found to occur in extremes of physical conditions not tolerated by any other chiton species, such as the extremely low salinity of the Baltic (Kaas & Van Belle, Reference Kaas and Van Belle1985). If a part of the physiological pathway is functioning close to its limit, there would be less capacity for the increases required in order to maintain uptake in depleted conditions. There may be some physiological trade-off between the higher activity rates that foraging in a resource-poor microhabitat requires, and the ability to regulate uptake in depleted conditions. The fact that L. cinerea is much more common and widespread than any other Atlantic chiton species suggests lack of oxyregulatory ability may not be a critical factor in its survival. Murdoch & Shumway (Reference Murdoch and Shumway1980) similarly found O. neglectus, the species operating at VO2 close to that observed here in L. cinerea, to have little oxyregulatory ability. However, they also observed C. porosus to be one of the weakest oxyregulators, but it had one of the lowest mass-specific VO2, suggesting that constraints caused by operating at higher metabolic rates are not solely responsible for a lack of oxyregulatory ability.
Phylogeny
There were no clear patterns in physiology correlated to membership of the two major polyplacophoran clades: Lepidopleurida and Chitonida. The term ‘living fossil’ is often applied to groups (like Polyplacophora) that have a deep fossil record and constrained morphology in extant taxa (Sirenko, Reference Sirenko2006). Although most chitons are superficially very similar in their body morphology, the extant diversity represents two separate radiations; the more recently derived order Chitonida (represented here by A. crinita, L. cinerea, M. ferreirai and T. lineata) and the more ‘primitive’ order Lepidopleurida (represented here by two Leptochiton spp.) (Sirenko, Reference Sirenko2006). The two clades can be distinguished anatomically by features of the gill arrangement, with Lepidopleurida typically having fewer gills (Yonge, Reference Yonge1939; Sirenko, Reference Sirenko1993; Sigwart, Reference Sigwart2008). Lepidopleurida are often found in the deep sea and are almost exclusively subtidal, while species occurring in the intertidal are almost always in Chitonida (Sigwart et al., Reference Sigwart, Schwabe, Saito, Samadi and Giribet2011). Both of the ‘primitive’ Leptochiton species studied here demonstrated clear oxyconforming behaviour, suggesting lepidopleuran species may lack adaptations to oxyregulation possessed by some members of Chitonida. A lack of oxyregulatory ability may have constrained lepidopleuran chitons from colonizing the dynamic intertidal habitat where fluxes in oxygen availability are common (Eernisse & Reynolds, Reference Eernisse, Reynolds, Harrison and Kohn1994). However, this has not prevented L. cinerea (in the more recently derived Chitonida) from extensively colonizing the intertidal and being found at the highest shore level of any north-east Atlantic chiton.
Conclusions
Factors such as microhabitat structure, body mass and phylogenetic position are among the main drivers that collectively underpin differences in physiological traits between species. The physiological variability observed here and in earlier studies (e.g. Murdoch & Shumway, Reference Murdoch and Shumway1980) is likely to be common across other taxa, and demonstrates how basal physiological traits can show significant complexity even within a highly morphologically constrained group. Such physiological variability has implications for studies of the ecology of such communities (Padilla & Allen, Reference Padilla and Allen2000).
In marine ecology, morphologically similar, co-occurring groups of species such as chitons may often be reported within single functional or taxonomic units without further detail (Steneck & Watling, Reference Steneck and Watling1982; Ortiz & Wolff, Reference Ortiz and Wolff2002; Bishop, Reference Bishop2003; Hetherington & Reid, Reference Hetherington and Reid2003). The predicted impacts of environmental change, such as changes in physical conditions, may be presumed to have equal influence on species grouped together within these guilds. Variation in physiology among commonly amalgamated taxa, such as those observed here, means this is unlikely to be the case, and impacts may be unpredictable. Another implication is that the full role or contribution of component groups to ecosystem functioning may be significantly understated (Padilla & Allen, Reference Padilla and Allen2000; Boero et al., Reference Boero, Bouillon, Gravili, Miglietta, Parsons and Piraino2008). While studies of marine ecology necessarily involve a level of abstraction of complex real world systems into more easily studied guilds, these findings demonstrate that such research must give due consideration to the extent of intra-guild variation at a finer taxonomic level.
ACKNOWLEDGEMENTS
The authors are grateful to Jose María Fariñas and Emma Gorman of Queen's University Marine Laboratory for assistance in collection of Leptochiton asellus specimens, and Lisa Kirkendale (University of Wollongong) for assistance collecting Leptochiton rugatus in British Columbia; Rob Mrowicki (Queen's University Belfast) provided a photograph in Figure 1; Enrico Schwabe (Zoologische Staatssammlung, Munich) helped with access to rare literature; J.D.R. Houghton (Queen's University Belfast) gave constructive comments about the manuscript. Douglas Eernisse and two anonymous referees substantially improved an earlier version of this manuscript. N.C. is supported by a PhD scholarship from the Department of Education and Learning, Northern Ireland. J.D.S. acknowledges support from University College Dublin for travel to BC, and J.G.R. acknowledges support from the Natural Sciences and Engineering Research Council of Canada.
