Journal of Fish Biology (2015) 86, 1680–1698 doi:10.1111/jfb.12672, available online at wileyonlinelibrary.com
Contribution of water chemistry and fish condition to otolith chemistry: comparisons across salinity environments C. Izzo*, Z. A. Doubleday, A. G. Schultz, S. H. Woodcock and B. M. Gillanders Southern Seas Ecology Laboratories, School of Biological Sciences, The University of Adelaide, Adelaide, SA 5005, Australia (Received 1 July 2014, Accepted 24 February 2015) This study quantified the per cent contribution of water chemistry to otolith chemistry using enriched stable isotopes of strontium (86 Sr) and barium (137 Ba). Euryhaline barramundi Lates calcarifer, were reared in marine (salinity 40), estuarine (salinity 20) and freshwater (salinity 0) under different temperature treatments. To calculate the contribution of water to Sr and Ba in otoliths, enriched isotopes in the tank water and otoliths were quantified and fitted to isotope mixing models. Fulton’s K and RNA:DNA were also measured to explore the influence of fish condition on sources of element uptake. Water was the predominant source of otolith Sr (between 65 and 99%) and Ba (between 64 and 89%) in all treatments, but contributions varied with temperature (for Ba), or interactively with temperature and salinity (for Sr). Fish condition indices were affected independently by the experimental rearing conditions, as RNA:DNA differed significantly among salinity treatments and Fulton’s K was significantly different between temperature treatments. Regression analyses did not detect relations between fish condition and per cent contribution values. General linear models indicated that contributions from water chemistry to otolith chemistry were primarily influenced by temperature and secondly by fish condition, with a relatively minor influence of salinity. These results further the understanding of factors that affect otolith element uptake, highlighting the necessity to consider the influence of environment and fish condition when interpreting otolith element data to reconstruct the environmental histories of fish. © 2015 The Fisheries Society of the British Isles
Key words: Ba isotopes; otolith element uptake; salinity; Sr isotopes; temperature.
INTRODUCTION The chemical analysis of otoliths has become a valuable tool for reconstructing the environmental histories of teleosts (Elsdon & Gillanders, 2003a; Gillanders, 2005; Elsdon et al., 2008). Water and diet are two sources of elements in the otolith (Campana, 1999), with water largely identified as the predominant source in species tested to date (Doubleday et al., 2013). The contribution of water and diet to otolith chemistry can, however, be influenced by environmental factors, such as temperature and salinity (Webb et al., 2012). Environmental factors directly affect physiological processes, such as osmoregulation, respiration, metabolism and growth (Buckley et al., 1999; Bœuf & Payan, 2001), and thus the rate of elemental uptake into the otolith (Campana, 1999). *Author to whom correspondence should be addressed: Tel.: +61 8 8313 7036; email: [email protected]
1680 © 2015 The Fisheries Society of the British Isles
C O N T R I B U T I O N O F WAT E R T O O T O L I T H C H E M I S T RY
1681
Osmotic regulation is required by all teleosts, irrespective of their external salinity environment (McCormick, 2001). Homeostasis in marine fishes is achieved by drinking seawater, introducing elements into the blood plasma via the intestine. Conversely, in freshwater fishes, osmoregulation and elemental uptake occur via the gills (Campana, 1999; McCormick, 2001). As regulatory pathways of elemental uptake are likely to differ between the gill and intestinal interface (Melancon et al., 2009), the relative contribution of water to otolith chemistry may also differ among freshwater teleosts and their marine counterparts (Wells et al., 2003; Elsdon et al., 2008). This renders direct comparisons of element deposition among obligate freshwater and marine species tenuous (Doubleday et al., 2013), impeding assessments of the relative influence of salinity environment on the elemental otolith uptake (Zimmerman, 2005). This is particularly problematic as otolith element data are commonly used to reconstruct the migratory patterns of fishes between salinity environments (Milton et al., 2000; Elsdon and Gillanders, 2003a; Gillanders, 2005). Increased ambient water temperature necessitates an increase in food consumption to meet greater metabolic demands (Bœuf & Payan, 2001; Katersky & Carter, 2007a, b), which in turn increases the availability of dietary sourced elements for incorporation (Webb et al., 2012). Furthermore, increased rates of somatic (and otolith) growth at higher temperatures exert kinetic and physiological controls on otolith element incorporation, which likely varies among species and life-history stage (Walther et al., 2010a). Altogether, these temperature-mediated processes have the potential to increase the uptake of dietary sourced elements into otoliths, reducing the relative contributions of water to otolith chemistry. Consideration of condition indices in relation to environmental variables, which are inherently linked, provides a means of further understanding the influences of temperature and salinity on sources of otolith element uptake. Fluctuations in ambient water temperature and salinity have been shown to alter the growth and condition of fishes (Katersky & Carter, 2007b; Domingos et al., 2013), as well as modifying the chemical composition of otoliths (Milton et al., 2008). This is particularly relevant for diadromous species that utilize both marine and freshwater habitats as transitions between salinity environments are likely to alter the growth and condition of the fish due to changes in osmotic regulatory demands (Bœuf & Payan, 2001). While many studies have investigated diadromous migratory patterns using otolith chemistry (Gillanders, 2005), few studies have explored the relative influence of fish condition on elemental otolith uptake, which may have an equal or greater influence on otolith chemistry and may in turn affect the accuracy of interpreting the environmental histories of individual fishes (Sturrock et al., 2012; Walther & Limburg, 2012). This study examined the effects of temperature and salinity on the per cent contribution of water chemistry to otolith strontium (Sr) and barium (Ba) in barramundi Lates calcarifer (Bloch 1790). Lates calcarifer is catadromous in the wild (Milton et al., 2000), providing an opportunity to directly compare the effects of different salinity environments (i.e. marine, estuarine and fresh water) to contributions of water to otolith chemistry. This study also sought to better understand how the uptake of elements from the water into otoliths is further influenced by short (RNA:DNA) and long-term (Fulton’s K) indices of condition.
© 2015 The Fisheries Society of the British Isles, Journal of Fish Biology 2015, 86, 1680–1698
1682
C . I Z Z O E T A L.
MATERIALS AND METHODS E X P E R I M E N TA L D E S I G N Lates calcarifer fingerlings (c. 10 weeks of age) were obtained from the Robarra Hatchery (www.robarra.com.au) at West Beach and delivered to the University of Adelaide, South Australia. Fingerlings had been raised at 27∘ C at a salinity of 10. Upon arrival, fingerlings were held in a 200 l holding tank filled with dilute seawater (salinity c. 10). Fingerlings were fed on a marine-based 1⋅5 mm floating pellet diet (Lucky Star; www.luckystarfeed.net) throughout the pre and experimental periods. Five days after arrival, fingerlings were batch marked via immersion in a 40 l tank filled with an alizarin complexone (C19 H15 NO8 ) solution (40 mg l−1 in bore water) for 24 h (van der Walt & Faragher, 2003), as a way of distinguishing between experimental and hatchery growth. Following marking, fingerlings were randomly distributed among experimental tanks at a density of 10–15 fish per tank. Fish were gradually acclimatized to experimental temperatures at a rate of 2∘ C per 24 h, using external heating units to increase temperature. Acclimation to the different salinity treatments was achieved by raising or lowering salinities at a rate of 5 every 24 h. Each tank contained 40 l of water, along with a small submerged filter and aerator. The overall experimental design consisted of three treatment variables: temperature, salinity and isotope enrichment. Water temperature was maintained at two treatment levels: ambient room temperature (c. 26∘ C) and heated to 30∘ C. Each temperature treatment level had 12 tanks which were randomly distributed across three water baths (n = 24 experimental tanks in total). Within each temperature treatment, tanks were separated into three salinity levels: marine (salinity 40), estuarine (salinity 20) and fresh water (salinity 0). Seawater (salinity 40) was sourced from the South Australian Research and Development Institute at West Beach, South Australia. Water in the freshwater tanks was aged (dechlorinated) tap water that had been aerated for >24 h in a reservoir tank. The estuarine water consisted of 1:1 seawater:freshwater mix. For each combination of temperature and salinity, the four tanks were further split into two isotope treatments: water enriched with a combination of 137 Ba and 86 Sr isotopes and a non-enriched control. There were two replicate tanks for each treatment combination and 10–15 replicate fish within each tank. Water was enriched by dissolving isotopically enriched BaCO3 or SrCO3 (Oak Ridge National Laboratories; www.orni.gov) at a concentration of 0⋅03 mg l−1 for 137 Ba and 0⋅25 mg l−1 for 86 Sr. Enrichment concentrations were based on baseline levels of Sr and Ba in the marine water (Sr = 85⋅93 mg l−1 and Ba = 0⋅08 mg l−1 ). Although baseline concentrations of Sr and Ba varied among the three water types, a single isotope enrichment concentration was used for all salinity treatments, ensuring a constant isotope enrichment volume (Webb et al., 2012). Fish were fed twice daily to satiation and any excess food was removed from all tanks 1 h after feeding. Water temperature and salinity were measured twice weekly throughout the experiment. All tanks had weekly 25% water changes, and tanks were topped up between changes to maintain salinity and isotope enrichment due to evaporation. Additional water isotope enrichment was achieved when water was replenished (not when topped up). Ammonia levels were tested regularly to ensure that water quality was maintained. Fish were exposed to experimental conditions for up to 34 days. At the conclusion of the experiment, all of the fish were euthanized in ice slurry and a portion of white pectoral muscle tissue was removed and frozen at −80∘ C for nucleic acid quantification. Standard lengths (LS ± mm) and wet body mass (W B ± gm) were recorded. Fulton’s K indices were based on LS and W B measurements (Mommsen, 1998). N U C L E I C A C I D Q U A N T I F I C AT I O N Frozen tissue samples were lysed in a tris-EDTA buffering solution containing 1% N-lauroylsarcosine (Caldarone et al., 2001) and the genomic supernatant was collected for RNA:DNA analysis. Nucleic acid quantification was performed using a Qubit 2.0 bench top fluorometer and the associated DNA and RNA assay kits following prescribed methods (Invitrogen; wwwlifetechnologies.com). Standard curves were calculated using the Qubit fluorometer’s software, based on analysis of supplied reference standards. All nucleic acid
© 2015 The Fisheries Society of the British Isles, Journal of Fish Biology 2015, 86, 1680–1698
C O N T R I B U T I O N O F WAT E R T O O T O L I T H C H E M I S T RY
1683
Table I. Operating parameters for the Agilent 7500cs inductively coupled plasma mass spectrometer (ICP-MS) and the New Wave UP213 laser ablation system with ICP-MS Solution ICP-MS Collision cell Cone Integration time Laser Wave length Mode Frequency Spot size Laser power Carrier Laser ICP-MS Optional gas Cone Dwell times (in ms)
He (5 ml min−1 ) Pt 0⋅1 s with three replicates for each isotope 213 nm Q-switch 5 Hz 30 μm 80% Ar (0⋅92 l min−1 ) He (58%) Pt 138 Ba (300), 137 Ba (400), 86 Sr and 88 Sr (200), 43 Ca (100) and 115 In (50)
quantifications were undertaken within a single session. For each specimen, replicate aliquots of the genomic extracts were analysed to quantify the DNA and RNA component. The c.v. of the repeated measures of DNA and RNA concentrations were low (c.v.% = 1⋅48 and 0⋅69, respectively). A sample containing no genomic material was also analysed periodically to test for potential contamination.
WAT E R I S O T O P E A N A LY S E S Water samples (25 ml) were collected from each tank at the beginning, middle and conclusion of the experiment. Each sample was filtered through a 0⋅45 μm filter and acidified with ultrapure nitric acid to create a 2% HNO3 solution and frozen until analysis. For the analysis of Sr and Ba isotopes, samples from the control freshwater tanks were analysed undiluted, while all other treatments were diluted 10:1, due to differing ambient elemental concentrations between treatments. Water samples were analysed using an Agilent 7500cs inductively coupled plasma mass spectrometer (ICP-MS; www.chem.agilent.com) (see Table I for operating parameters). All samples were analysed for 86 Sr, 88 Sr, 137 Ba, 138 Ba and 43 Ca for element Ca ratios. Agilent Mass Hunter software was used to collect the raw data, which were calibrated against a multi-element standard. The elemental standard was used to measure instrument drift and precision, which was deemed to be acceptable (c.v. < 5%). An indium spike was also used to measure element recoveries, which were deemed to be acceptable (recoveries of 98–102%). Water element concentration data were adjusted using a correction factor based on deviations from expected natural isotopic ratios in the control treatment tanks. O T O L I T H P R E PA R AT I O N A N D I S O T O P E A N A LY S E S Sagittal otoliths were removed and cleaned of any adhering tissue before being air-dried. One otolith from each fish was embedded in a clear setting epoxy resin (EpoFix; www.struers.com), which was spiked with 40 mg l−1 indium to enable discrimination between the otolith and resin during analysis. Embedded otoliths were then thin-sectioned (c. 500 μm) using a low-speed saw (Buehler; www.buehler.com) and polished using progressively finer grades of lapping film. Polished sections were mounted onto a microscope slide using indium-spiked thermoplastic glue (Crystalbond; www.crystalbond.com).
© 2015 The Fisheries Society of the British Isles, Journal of Fish Biology 2015, 86, 1680–1698
1684
C . I Z Z O E T A L.
Otoliths were analysed using a New Wave UP213 laser ablation system connected to an Agilent 7500cs ICP-MS (www.agilent.com) (see Table I for operating parameters). Spot ablations were used to sample the marginal edge of the otolith, which represented experimental growth only (between the alizarin mark and the otolith edge). All samples were analysed for 86 Sr, 88 Sr, 137 Ba, 138 Ba and 43 Ca to determine elemental ratios, as well as 115 In as an indicator of when the laser was ablating epoxy or thermoplastic glue rather than otolith material. After every 10th sample, a glass reference standard (NIST612) was analysed and used to correct for mass bias and machine drift. Similarly, at the beginning and conclusion of each analysis session, an internal CaCO3 standard (MACS-3: United States Geological Survey; http://crustal.usgs.gov/geochemical_reference_standards/microanalytical_RM.html/) was analysed as a measure of precision, which was deemed to be acceptable (c.v. < 6%). Otolith element concentration data were adjusted using a correction factor based on the deviation of the raw concentration data from the known isotopic concentration of the MACS-3 carbonate standard. The Glitter software package was used to collect the raw data (Griffin et al., 2008). Further data reduction was undertaken in Excel (Microsoft) and all isotope data were smoothed using a six-point running mean and the mean of the smoothed data used as spot depth profiles displayed an overall consistent signature for the experimental portion analysed. To calculate total Ba:Ca and Sr:Ca (in mmol mol−1 ), 138 Ba and 88 Sr isotopes were ratioed to 43 Ca.
P E R C E N T C O N T R I B U T I O N C A L C U L AT I O N S The per cent contributions of the isotope-enriched water to otolith Sr and Ba was calculated using a linear isotope mixing model for individual fish (Kennedy et al., 2000): % Cw = 100{1 − [(Rw − Ro )(Rw − Rd )− 1 ]}, where R is the isotope ratio of interest and %Cw is the per cent contribution of water to the otolith, and the subscripts o, w and d represent the otolith, water and diet. The isotopic ratios of Ba and Sr measured in the control tanks were used for the water component of the equation. The natural isotope values for the diet (88 Sr:86 Sr = 8⋅38 and 138 Ba:137 Ba = 6⋅38) were used for the dietary component of the equation. S TAT I S T I C A L A N A LY S E S Statistical analysis was conducted using Primer 6.0 (Clarke & Gorley, 2006). For all tests, non-transformed data were fitted to Euclidean distance matrices and performed using unrestricted permutations of the data. Three-factor and four-factor ANOVA was used to compare Fulton’s K and RNA:DNA among treatments, as well as to test for differences in isotope (88 Sr:86 Sr and 138 Ba:137 Ba) and elemental (Ba:Ca and Sr:Ca) ratios in both water samples and otoliths, analysing for the effects of temperature, salinity and isotope treatment. Similar tests were performed on the per cent contributions of both the enriched Sr and Ba otolith chemistry. Additional tests were performed to determine variations in rearing conditions among tanks. Temperature, salinity and isotope treatment were treated as fixed factors. For fish condition indices and water data, tank was treated as a random factor and nested within the other three variables. Despite differences in rearing conditions (Table SI, Supporting Information), there were no significant differences found in otolith chemistry among tanks, and thus data for replicate tanks were pooled and re-analysed without the nested tank term. If significant differences occurred, then post hoc pair-wise tests were conducted to determine where differences occurred. Regressions were used to determine whether Fulton’s K and RNA:DNA were significantly related to one another, as well as element:Ca and per cent contribution values for both Sr and Ba. The influences of fish condition and environmental factors on the contribution of water chemistry to otolith Ba and Sr were assessed using generalized linear models. All combinations of temperature and salinity as well as the condition indices, Fulton’s K and RNA:DNA, were sequentially added into an explanatory model set (n = 10). All models were ranked using Akaike’s information criterion (AIC) corrected for small sample sizes (AICc ) (Burnham & Anderson, 2004). The ratio of evidence for the highest ranked model was calculated by dividing the weighed AICc (wAICc ) for the best model by the wAICc of one of the two base models that either included the terms temperature or salinity. If the difference between the second ranked model’s ΔAICc value and that of the best model was
Contribution of water chemistry and fish condition to otolith chemistry: comparisons across salinity environments C. Izzo*, Z. A. Doubleday, A. G. Schultz, S. H. Woodcock and B. M. Gillanders Southern Seas Ecology Laboratories, School of Biological Sciences, The University of Adelaide, Adelaide, SA 5005, Australia (Received 1 July 2014, Accepted 24 February 2015) This study quantified the per cent contribution of water chemistry to otolith chemistry using enriched stable isotopes of strontium (86 Sr) and barium (137 Ba). Euryhaline barramundi Lates calcarifer, were reared in marine (salinity 40), estuarine (salinity 20) and freshwater (salinity 0) under different temperature treatments. To calculate the contribution of water to Sr and Ba in otoliths, enriched isotopes in the tank water and otoliths were quantified and fitted to isotope mixing models. Fulton’s K and RNA:DNA were also measured to explore the influence of fish condition on sources of element uptake. Water was the predominant source of otolith Sr (between 65 and 99%) and Ba (between 64 and 89%) in all treatments, but contributions varied with temperature (for Ba), or interactively with temperature and salinity (for Sr). Fish condition indices were affected independently by the experimental rearing conditions, as RNA:DNA differed significantly among salinity treatments and Fulton’s K was significantly different between temperature treatments. Regression analyses did not detect relations between fish condition and per cent contribution values. General linear models indicated that contributions from water chemistry to otolith chemistry were primarily influenced by temperature and secondly by fish condition, with a relatively minor influence of salinity. These results further the understanding of factors that affect otolith element uptake, highlighting the necessity to consider the influence of environment and fish condition when interpreting otolith element data to reconstruct the environmental histories of fish. © 2015 The Fisheries Society of the British Isles
Key words: Ba isotopes; otolith element uptake; salinity; Sr isotopes; temperature.
INTRODUCTION The chemical analysis of otoliths has become a valuable tool for reconstructing the environmental histories of teleosts (Elsdon & Gillanders, 2003a; Gillanders, 2005; Elsdon et al., 2008). Water and diet are two sources of elements in the otolith (Campana, 1999), with water largely identified as the predominant source in species tested to date (Doubleday et al., 2013). The contribution of water and diet to otolith chemistry can, however, be influenced by environmental factors, such as temperature and salinity (Webb et al., 2012). Environmental factors directly affect physiological processes, such as osmoregulation, respiration, metabolism and growth (Buckley et al., 1999; Bœuf & Payan, 2001), and thus the rate of elemental uptake into the otolith (Campana, 1999). *Author to whom correspondence should be addressed: Tel.: +61 8 8313 7036; email: [email protected]
1680 © 2015 The Fisheries Society of the British Isles
C O N T R I B U T I O N O F WAT E R T O O T O L I T H C H E M I S T RY
1681
Osmotic regulation is required by all teleosts, irrespective of their external salinity environment (McCormick, 2001). Homeostasis in marine fishes is achieved by drinking seawater, introducing elements into the blood plasma via the intestine. Conversely, in freshwater fishes, osmoregulation and elemental uptake occur via the gills (Campana, 1999; McCormick, 2001). As regulatory pathways of elemental uptake are likely to differ between the gill and intestinal interface (Melancon et al., 2009), the relative contribution of water to otolith chemistry may also differ among freshwater teleosts and their marine counterparts (Wells et al., 2003; Elsdon et al., 2008). This renders direct comparisons of element deposition among obligate freshwater and marine species tenuous (Doubleday et al., 2013), impeding assessments of the relative influence of salinity environment on the elemental otolith uptake (Zimmerman, 2005). This is particularly problematic as otolith element data are commonly used to reconstruct the migratory patterns of fishes between salinity environments (Milton et al., 2000; Elsdon and Gillanders, 2003a; Gillanders, 2005). Increased ambient water temperature necessitates an increase in food consumption to meet greater metabolic demands (Bœuf & Payan, 2001; Katersky & Carter, 2007a, b), which in turn increases the availability of dietary sourced elements for incorporation (Webb et al., 2012). Furthermore, increased rates of somatic (and otolith) growth at higher temperatures exert kinetic and physiological controls on otolith element incorporation, which likely varies among species and life-history stage (Walther et al., 2010a). Altogether, these temperature-mediated processes have the potential to increase the uptake of dietary sourced elements into otoliths, reducing the relative contributions of water to otolith chemistry. Consideration of condition indices in relation to environmental variables, which are inherently linked, provides a means of further understanding the influences of temperature and salinity on sources of otolith element uptake. Fluctuations in ambient water temperature and salinity have been shown to alter the growth and condition of fishes (Katersky & Carter, 2007b; Domingos et al., 2013), as well as modifying the chemical composition of otoliths (Milton et al., 2008). This is particularly relevant for diadromous species that utilize both marine and freshwater habitats as transitions between salinity environments are likely to alter the growth and condition of the fish due to changes in osmotic regulatory demands (Bœuf & Payan, 2001). While many studies have investigated diadromous migratory patterns using otolith chemistry (Gillanders, 2005), few studies have explored the relative influence of fish condition on elemental otolith uptake, which may have an equal or greater influence on otolith chemistry and may in turn affect the accuracy of interpreting the environmental histories of individual fishes (Sturrock et al., 2012; Walther & Limburg, 2012). This study examined the effects of temperature and salinity on the per cent contribution of water chemistry to otolith strontium (Sr) and barium (Ba) in barramundi Lates calcarifer (Bloch 1790). Lates calcarifer is catadromous in the wild (Milton et al., 2000), providing an opportunity to directly compare the effects of different salinity environments (i.e. marine, estuarine and fresh water) to contributions of water to otolith chemistry. This study also sought to better understand how the uptake of elements from the water into otoliths is further influenced by short (RNA:DNA) and long-term (Fulton’s K) indices of condition.
© 2015 The Fisheries Society of the British Isles, Journal of Fish Biology 2015, 86, 1680–1698
1682
C . I Z Z O E T A L.
MATERIALS AND METHODS E X P E R I M E N TA L D E S I G N Lates calcarifer fingerlings (c. 10 weeks of age) were obtained from the Robarra Hatchery (www.robarra.com.au) at West Beach and delivered to the University of Adelaide, South Australia. Fingerlings had been raised at 27∘ C at a salinity of 10. Upon arrival, fingerlings were held in a 200 l holding tank filled with dilute seawater (salinity c. 10). Fingerlings were fed on a marine-based 1⋅5 mm floating pellet diet (Lucky Star; www.luckystarfeed.net) throughout the pre and experimental periods. Five days after arrival, fingerlings were batch marked via immersion in a 40 l tank filled with an alizarin complexone (C19 H15 NO8 ) solution (40 mg l−1 in bore water) for 24 h (van der Walt & Faragher, 2003), as a way of distinguishing between experimental and hatchery growth. Following marking, fingerlings were randomly distributed among experimental tanks at a density of 10–15 fish per tank. Fish were gradually acclimatized to experimental temperatures at a rate of 2∘ C per 24 h, using external heating units to increase temperature. Acclimation to the different salinity treatments was achieved by raising or lowering salinities at a rate of 5 every 24 h. Each tank contained 40 l of water, along with a small submerged filter and aerator. The overall experimental design consisted of three treatment variables: temperature, salinity and isotope enrichment. Water temperature was maintained at two treatment levels: ambient room temperature (c. 26∘ C) and heated to 30∘ C. Each temperature treatment level had 12 tanks which were randomly distributed across three water baths (n = 24 experimental tanks in total). Within each temperature treatment, tanks were separated into three salinity levels: marine (salinity 40), estuarine (salinity 20) and fresh water (salinity 0). Seawater (salinity 40) was sourced from the South Australian Research and Development Institute at West Beach, South Australia. Water in the freshwater tanks was aged (dechlorinated) tap water that had been aerated for >24 h in a reservoir tank. The estuarine water consisted of 1:1 seawater:freshwater mix. For each combination of temperature and salinity, the four tanks were further split into two isotope treatments: water enriched with a combination of 137 Ba and 86 Sr isotopes and a non-enriched control. There were two replicate tanks for each treatment combination and 10–15 replicate fish within each tank. Water was enriched by dissolving isotopically enriched BaCO3 or SrCO3 (Oak Ridge National Laboratories; www.orni.gov) at a concentration of 0⋅03 mg l−1 for 137 Ba and 0⋅25 mg l−1 for 86 Sr. Enrichment concentrations were based on baseline levels of Sr and Ba in the marine water (Sr = 85⋅93 mg l−1 and Ba = 0⋅08 mg l−1 ). Although baseline concentrations of Sr and Ba varied among the three water types, a single isotope enrichment concentration was used for all salinity treatments, ensuring a constant isotope enrichment volume (Webb et al., 2012). Fish were fed twice daily to satiation and any excess food was removed from all tanks 1 h after feeding. Water temperature and salinity were measured twice weekly throughout the experiment. All tanks had weekly 25% water changes, and tanks were topped up between changes to maintain salinity and isotope enrichment due to evaporation. Additional water isotope enrichment was achieved when water was replenished (not when topped up). Ammonia levels were tested regularly to ensure that water quality was maintained. Fish were exposed to experimental conditions for up to 34 days. At the conclusion of the experiment, all of the fish were euthanized in ice slurry and a portion of white pectoral muscle tissue was removed and frozen at −80∘ C for nucleic acid quantification. Standard lengths (LS ± mm) and wet body mass (W B ± gm) were recorded. Fulton’s K indices were based on LS and W B measurements (Mommsen, 1998). N U C L E I C A C I D Q U A N T I F I C AT I O N Frozen tissue samples were lysed in a tris-EDTA buffering solution containing 1% N-lauroylsarcosine (Caldarone et al., 2001) and the genomic supernatant was collected for RNA:DNA analysis. Nucleic acid quantification was performed using a Qubit 2.0 bench top fluorometer and the associated DNA and RNA assay kits following prescribed methods (Invitrogen; wwwlifetechnologies.com). Standard curves were calculated using the Qubit fluorometer’s software, based on analysis of supplied reference standards. All nucleic acid
© 2015 The Fisheries Society of the British Isles, Journal of Fish Biology 2015, 86, 1680–1698
C O N T R I B U T I O N O F WAT E R T O O T O L I T H C H E M I S T RY
1683
Table I. Operating parameters for the Agilent 7500cs inductively coupled plasma mass spectrometer (ICP-MS) and the New Wave UP213 laser ablation system with ICP-MS Solution ICP-MS Collision cell Cone Integration time Laser Wave length Mode Frequency Spot size Laser power Carrier Laser ICP-MS Optional gas Cone Dwell times (in ms)
He (5 ml min−1 ) Pt 0⋅1 s with three replicates for each isotope 213 nm Q-switch 5 Hz 30 μm 80% Ar (0⋅92 l min−1 ) He (58%) Pt 138 Ba (300), 137 Ba (400), 86 Sr and 88 Sr (200), 43 Ca (100) and 115 In (50)
quantifications were undertaken within a single session. For each specimen, replicate aliquots of the genomic extracts were analysed to quantify the DNA and RNA component. The c.v. of the repeated measures of DNA and RNA concentrations were low (c.v.% = 1⋅48 and 0⋅69, respectively). A sample containing no genomic material was also analysed periodically to test for potential contamination.
WAT E R I S O T O P E A N A LY S E S Water samples (25 ml) were collected from each tank at the beginning, middle and conclusion of the experiment. Each sample was filtered through a 0⋅45 μm filter and acidified with ultrapure nitric acid to create a 2% HNO3 solution and frozen until analysis. For the analysis of Sr and Ba isotopes, samples from the control freshwater tanks were analysed undiluted, while all other treatments were diluted 10:1, due to differing ambient elemental concentrations between treatments. Water samples were analysed using an Agilent 7500cs inductively coupled plasma mass spectrometer (ICP-MS; www.chem.agilent.com) (see Table I for operating parameters). All samples were analysed for 86 Sr, 88 Sr, 137 Ba, 138 Ba and 43 Ca for element Ca ratios. Agilent Mass Hunter software was used to collect the raw data, which were calibrated against a multi-element standard. The elemental standard was used to measure instrument drift and precision, which was deemed to be acceptable (c.v. < 5%). An indium spike was also used to measure element recoveries, which were deemed to be acceptable (recoveries of 98–102%). Water element concentration data were adjusted using a correction factor based on deviations from expected natural isotopic ratios in the control treatment tanks. O T O L I T H P R E PA R AT I O N A N D I S O T O P E A N A LY S E S Sagittal otoliths were removed and cleaned of any adhering tissue before being air-dried. One otolith from each fish was embedded in a clear setting epoxy resin (EpoFix; www.struers.com), which was spiked with 40 mg l−1 indium to enable discrimination between the otolith and resin during analysis. Embedded otoliths were then thin-sectioned (c. 500 μm) using a low-speed saw (Buehler; www.buehler.com) and polished using progressively finer grades of lapping film. Polished sections were mounted onto a microscope slide using indium-spiked thermoplastic glue (Crystalbond; www.crystalbond.com).
© 2015 The Fisheries Society of the British Isles, Journal of Fish Biology 2015, 86, 1680–1698
1684
C . I Z Z O E T A L.
Otoliths were analysed using a New Wave UP213 laser ablation system connected to an Agilent 7500cs ICP-MS (www.agilent.com) (see Table I for operating parameters). Spot ablations were used to sample the marginal edge of the otolith, which represented experimental growth only (between the alizarin mark and the otolith edge). All samples were analysed for 86 Sr, 88 Sr, 137 Ba, 138 Ba and 43 Ca to determine elemental ratios, as well as 115 In as an indicator of when the laser was ablating epoxy or thermoplastic glue rather than otolith material. After every 10th sample, a glass reference standard (NIST612) was analysed and used to correct for mass bias and machine drift. Similarly, at the beginning and conclusion of each analysis session, an internal CaCO3 standard (MACS-3: United States Geological Survey; http://crustal.usgs.gov/geochemical_reference_standards/microanalytical_RM.html/) was analysed as a measure of precision, which was deemed to be acceptable (c.v. < 6%). Otolith element concentration data were adjusted using a correction factor based on the deviation of the raw concentration data from the known isotopic concentration of the MACS-3 carbonate standard. The Glitter software package was used to collect the raw data (Griffin et al., 2008). Further data reduction was undertaken in Excel (Microsoft) and all isotope data were smoothed using a six-point running mean and the mean of the smoothed data used as spot depth profiles displayed an overall consistent signature for the experimental portion analysed. To calculate total Ba:Ca and Sr:Ca (in mmol mol−1 ), 138 Ba and 88 Sr isotopes were ratioed to 43 Ca.
P E R C E N T C O N T R I B U T I O N C A L C U L AT I O N S The per cent contributions of the isotope-enriched water to otolith Sr and Ba was calculated using a linear isotope mixing model for individual fish (Kennedy et al., 2000): % Cw = 100{1 − [(Rw − Ro )(Rw − Rd )− 1 ]}, where R is the isotope ratio of interest and %Cw is the per cent contribution of water to the otolith, and the subscripts o, w and d represent the otolith, water and diet. The isotopic ratios of Ba and Sr measured in the control tanks were used for the water component of the equation. The natural isotope values for the diet (88 Sr:86 Sr = 8⋅38 and 138 Ba:137 Ba = 6⋅38) were used for the dietary component of the equation. S TAT I S T I C A L A N A LY S E S Statistical analysis was conducted using Primer 6.0 (Clarke & Gorley, 2006). For all tests, non-transformed data were fitted to Euclidean distance matrices and performed using unrestricted permutations of the data. Three-factor and four-factor ANOVA was used to compare Fulton’s K and RNA:DNA among treatments, as well as to test for differences in isotope (88 Sr:86 Sr and 138 Ba:137 Ba) and elemental (Ba:Ca and Sr:Ca) ratios in both water samples and otoliths, analysing for the effects of temperature, salinity and isotope treatment. Similar tests were performed on the per cent contributions of both the enriched Sr and Ba otolith chemistry. Additional tests were performed to determine variations in rearing conditions among tanks. Temperature, salinity and isotope treatment were treated as fixed factors. For fish condition indices and water data, tank was treated as a random factor and nested within the other three variables. Despite differences in rearing conditions (Table SI, Supporting Information), there were no significant differences found in otolith chemistry among tanks, and thus data for replicate tanks were pooled and re-analysed without the nested tank term. If significant differences occurred, then post hoc pair-wise tests were conducted to determine where differences occurred. Regressions were used to determine whether Fulton’s K and RNA:DNA were significantly related to one another, as well as element:Ca and per cent contribution values for both Sr and Ba. The influences of fish condition and environmental factors on the contribution of water chemistry to otolith Ba and Sr were assessed using generalized linear models. All combinations of temperature and salinity as well as the condition indices, Fulton’s K and RNA:DNA, were sequentially added into an explanatory model set (n = 10). All models were ranked using Akaike’s information criterion (AIC) corrected for small sample sizes (AICc ) (Burnham & Anderson, 2004). The ratio of evidence for the highest ranked model was calculated by dividing the weighed AICc (wAICc ) for the best model by the wAICc of one of the two base models that either included the terms temperature or salinity. If the difference between the second ranked model’s ΔAICc value and that of the best model was
