UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Oxygen-limited thermal tolerance in hybrid killfish, Fundulus heteroclitus Haakons, Kristen Leigh 2010

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
24-ubc_2010_spring_haakons_kristen.pdf [ 1.49MB ]
Metadata
JSON: 24-1.0069995.json
JSON-LD: 24-1.0069995-ld.json
RDF/XML (Pretty): 24-1.0069995-rdf.xml
RDF/JSON: 24-1.0069995-rdf.json
Turtle: 24-1.0069995-turtle.txt
N-Triples: 24-1.0069995-rdf-ntriples.txt
Original Record: 24-1.0069995-source.json
Full Text
24-1.0069995-fulltext.txt
Citation
24-1.0069995.ris

Full Text

OXYGEN-LIMITED THERMAL TOLERANCE IN HYBRID KILLIFISH, Fundulus heteroclitus  by   Kristen Leigh Haakons    B.Sc., The University of British Columbia, 2007     A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE   in   THE FACULTY OF GRADUATE STUDIES (Zoology)      THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  March, 2010     © Kristen Leigh Haakons, 2010 ii  ABSTRACT  The oxygen-limited thermal tolerance (OLTT) hypothesis suggests that an organism’s upper and lower thermal tolerance limits are linked to the point at which the organism cannot extract enough oxygen from its environment to meet its metabolic demands. Therefore, one conclusion that could be drawn from this hypothesis is that an organism’s metabolic rate (Mo2) should be inversely correlated with the maximum environmental temperature it can tolerate. This is because organisms with a higher demand for oxygen will be the first to experience a mismatch between oxygen supply and demand as the amount of dissolved oxygen in water decreases with increasing temperature. The validity of this hypothesis was tested using the common killifish, Fundulus heteroclitus, from a contact zone between the two subspecies: the Northern Fundulus heteroclitus macrolepidotus and the Southern Fundulus heteroclitus heteroclitus. I found that there was a significant inverse correlation (p=0.0192) between the critical thermal maximum (CTMax) and routine metabolic rate (RMR) after controlling for the weight of fish, which supports the OLTT hypothesis. The predictive power of this relationship, however, was low (R 2 =0.16). In addition, there was a significant effect of collection location on CTMax (p<0.001) with fish that were collected from a more southern marsh having higher thermal tolerance than fish from a more northern marsh within the contact zone.  I also examined the relationship between nuclear and mitochondrial genotypes and 3 physiological measurements in fish from the contact zone: RMR, CTMax, and critical oxygen tension (Pcrit). Using 7 nuclear microsatellite markers, a hybrid index was generated for each individual in order to assess nuclear genotype. I found no correlation between nuclear genotype and any of the physiological parameters measured. I did see a trend towards a significant correlation between mitochondrial genotype on RMR, however (p=0.0595). Fish with Northern iii  mitochondrial genotypes tended to have higher RMR’s than Southern fish, as was found in previous work on fish from the Northern and Southern extremes of the range. Together these data suggest that mitochondrial genotype may have some effect on metabolic rate and that metabolic rate has some relationship with thermal tolerance, but that these linkages are subtle. iv   TABLE OF CONTENTS ABSTRACT.............................................................................................................................. ii TABLE OF CONTENTS ........................................................................................................ iv LIST OF TABLES ................................................................................................................... v LIST OF FIGURES ................................................................................................................ vi LIST OF ABBREVIATIONS ............................................................................................... viii ACKNOWLEDGEMENTS .................................................................................................... ix INTRODUCTION .................................................................................................................... 1 Oxygen limitation of thermal tolerance ................................................................................... 1 Genetic influences on phenotype ............................................................................................ 4 Interactions between mitochondrial and nuclear genotypes ..................................................... 7 Killifish as an experimental model .......................................................................................... 9 Killifish habitat ..................................................................................................................... 10 Killifish genetic variation ..................................................................................................... 11 Killifish physiological variation ............................................................................................ 15 Evidence for cytonuclear incompatibilities in killifish ........................................................... 20 Thesis objectives .................................................................................................................. 20 MATERIALS AND METHODS ........................................................................................... 22 Whole-animal respirometry .................................................................................................. 23 Thermal tolerance ................................................................................................................. 24 Microsatellite analysis and hybrid index determination ......................................................... 26 Mitochondrial haplotype determination ................................................................................ 28 Statistics ............................................................................................................................... 28 RESULTS ............................................................................................................................... 30 Phenotype of New Jersey fish ............................................................................................... 34 The effects of mitochondrial genotype and collection location on phenotype ........................ 40 The effect of nuclear genotype and mismatch with mitochondrial genotype .......................... 44 Testing the correlation between CTMax and Mo2 in fish from the contact zone .................... 47 DISCUSSION ......................................................................................................................... 49 Physiology of New Jersey fish .............................................................................................. 50 Oxygen-limited thermal tolerance across the species range ................................................... 55 New Jersey fish .................................................................................................................... 57 Effect of collection location and mitochondrial genome on physiology ................................. 58 Effect of nuclear genotype on phenotype .............................................................................. 62 Cytonuclear incompatibility .................................................................................................. 63 Oxygen-limited thermal tolerance in hybrids? ....................................................................... 64 CONCLUSIONS .................................................................................................................... 66 BIBLIOGRAPHY .................................................................................................................. 67 APPENDIX ............................................................................................................................. 75  v  LIST OF TABLES Table 1. Master mix ingredients in volumes and concentrations for microsatellite multiplex PCR Reactions 1, 2, and 3 ................................................................................................................ 26 Table 2. Mass specific metabolic rates of Northern (n= 19), New Jersey (n=57), and Southern (n=16) fish acclimated to 20°C ................................................................................................. 36 Table 3. Mass-specific routine metabolic rates (mean ± SEM) for Red Bank fish (n=37), Metedeconk fish with Northern mitotypes (n=11) and Metedeconk fish with Southern mitotypes (n=12) acclimated to 20°C. ....................................................................................................... 41   vi  LIST OF FIGURES Figure 1A. Linear regression between the logarithm of wet weight and the logarithm of routine metabolic rate of all Northern (n=19), New Jersey (n=57) and Southern fish (n=16) (p < 0.001).  ................................................................................................................................................ 31 Figure 1B. Linear regression between the logarithm of wet weight and thermal tolerance (CTMax) of all Northern (n=17), New Jersey (n=54) and Southern fish (n=12) (p=0.298). ....... 32 Figure 1C. Linear regression between the logarithm of wet weight and critical oxygen tension (Pcrit) of all Northern (n=15), New Jersey (n=54) and Southern fish (n=14) (p=0.625). ........... 33 Figure 2. Mean residuals of log Mo2 against log wet weight (A), CTMax (B), and Pcrit (C) values for each population of common killifish when acclimated to 20°C. All data are expressed as mean ± SEM, and significant differences between populations are indicated by different letters (p<0.001 for all comparisons)................................................................................................... 35 Figure 3. Inverse correlation between the residuals of metabolic rate, Mo2, and thermal tolerance, CTMax, in common killifish acclimated to 20°C. (Northern fish n=19; New Jersey n=63; Southern fish n=16; Pearson correlation: p=0.00102, r= -0.354) ..................................... 37 Figure 4. Significant positive correlation between residuals of metabolic rate, Mo2 and critical oxygen tension, Pcrit in common killifish acclimated to 20°C. (Northern fish n=14; New Jersey n=55; Southern fish n=14; Pearson correlation: p=2.2x10 -7 , r=0.535) ....................................... 38 Figure 5. No significant correlation between critical oxygen tension, Pcrit, and thermal tolerance, CTMax, in common killifish acclimated to 20°C. (Northern fish n=13; New Jersey fish n=49; Southern fish n=10; Pearson correlation: p=0.666, r= -0.0517) ....................................... 39 Figure 6. Mean CTMax values for groups by mitochondrial genotype (A) and sample location (B). All data is presented as means ± SEM. Significant differences are represented by different letters in part B (p=0.003). (Metedeconk Northern n=11, Metedeconk Southern n=12, Red Bank n= 31) ...................................................................................................................................... 42 Figure 7. Mean residual of Mo2’s for groups by mitochondrial genotype (A) and sample location (B). All data is presented as means ± SEM. (Red Bank Northern, n=37; Metedeconk Northern, n=11; Metedeconk Southern, n=12).......................................................................................... 43 Figure 8. The insignificant correlations between residuals of Mo2 (A), CTMax (B), and Pcrit (C) and hybrid index of common killifish at 20°C. Metedeconk Northern n=11 (Mo2 and CTMax) and 8 (Pcrit); Metedeconk Southern n= 12; Red Bank Northern n= 35 (Mo2) and 31 (CTMax and Pcrit). ......................................................................................................................................... 45 vii  Figure 9. The range of residual Mo2 (A), CTMax (B), and Pcrit (C) values of fish from Red Bank and Metedeconk with Northern mitotypes and varying nuclear genotype (hybrid index) values. Individuals to the left of the graph are more similar to Southern parents, whereas individuals on the right are more similar to Northern parents........................................................................... 46 Figure 10. The significant inverse correlation between residuals of Mo2 and CTMax in fish from New Jersey (p=0.0192) acclimated to 20°C. (Red Bank Northern, n=31; Metedeconk Northern, n=11; Metedeconk Southern, n=12).......................................................................................... 48  viii  LIST OF ABBREVIATIONS ANOVA  analysis of variance ATP   adenosine triphosphate bp   base pairs °C   degrees Celsius COX   complex IV CTMax  critical thermal maximum DNA   deoxyribonucleic acid dNTPs   deoxynucleotide triphosphates ETC   electron transport chain g   gram h   hour Hb   haemoglobin H.I.   hybrid index Hr   hour hsp   heat-shock protein l   litre Ldh   lactate dehydrogenase LOE   loss of equilibrium Mdh   malate dehydrogenase min   minutes mM   millimolar Mo2   metabolic rate mtDNA  mitochondrial deoxyribonucleic acid O2   oxygen [O2]   concentration of oxygen OLTT   oxygen-limited thermal tolerance PCR   polymerase chain reaction Pcrit   critical oxygen tension PGI   phosphoglucose isomerise PGM   phosphoglucomutase Po2   partial pressure of oxygen Ppt   parts per thousand RMR   routine metabolic rate s   second SEM   standard error of the mean µl   microlitre ix  ACKNOWLEDGEMENTS  It would be impossible to sufficiently express my gratitude and admiration for the people who have helped me develop both personally and academically to this point in my education. I will start, however, by thanking my supervisor, Dr. Patricia Schulte whose guidance and tireless support have made this milestone in my life possible. Her breadth of knowledge and unfailing patience has made my time in graduate studies that much more enjoyable and her support and understanding through various personal problems are greatly appreciated. I am so grateful to have had the opportunity to work with this inspiring woman.  I would also like to thank my tireless committee, Drs. Darren Irwin and Jeffrey Richards, who have braved countless questions from me and always answered them with a smile on their faces, no matter how ridiculous my questions may have been. There are not enough words or space in these acknowledgements to say how thankful I am that I have had their brains to pick during this process. The people and the department of Zoology deserve special thanks as well for providing such an inspiring place to learn and do research.  Finally, I would like to thank my Mom, Dad, and Stepmom for having faith in me and encouraging me when times were rough. No matter how stressed out I was or how lazy I was feeling, they always seemed to know exactly what to say to get me back on track. Even though they may not understand a word of this thesis, they are proud to say I have done it and I can’t begin to describe how lucky I feel to have all of their support. 1  INTRODUCTION  Temperature is thought to play an important role in determining the distribution of organisms (Fry 1947; Hochachka and Somero 2002) due to the effects that temperature has on the rates of physiological and biochemical reactions (Hochachka and Somero 2002). Recently, research into the mechanisms by which temperature acts as a determinant of species biogeography has become increasingly prevalent due to growing awareness of the effects of human activities on global temperatures. Many studies have already demonstrated shifts in marine species distributions and community structure as a result of global warming (Perry et al. 2005; Grebmeier et al. 2006; Pörtner and Knust 2007). Understanding the mechanisms by which temperature acts to influence an organism’s distribution could allow us to make predictions about the shifts in population structure and the potential consequences that climate change will have on species.  The overall goal of my thesis research is to explore some potential mechanisms by which temperature may affect the physiological functions of animals. To do this I tested one of the predictions of a recent theory: the theory of “Oxygen-Limited Thermal Tolerance”, which predicts that the ability of an organism to withstand high temperatures should be negatively correlated with its routine metabolic rate. In addition, I addressed the question of whether variation in nuclear and mitochondrial genomes could account for variation in either metabolic rate or thermal tolerance. Oxygen limitation of thermal tolerance  One of the many limitations animals face in aquatic environments is the ability to extract enough oxygen from the water to match the demand dictated by the organism’s biochemical 2  processes. This challenge is exaggerated in intertidal marine organisms whose environments vary daily and seasonally in temperature, dissolved oxygen, pH, and salinity (Truchot and Duhamel-Jouve 1980). Temperature has been shown to affect cellular metabolic rates and thus, the organism’s oxygen demand, whereas dissolved oxygen levels in the environment dictate the amount of oxygen available to supply this metabolic demand (Fry 1947; Hughes 1973; Pörtner 2001; Fangue 2009a). Pörtner (2002) suggests that it is actually oxygen limitations which set organismal thermal tolerance limits at both high and low temperatures. He suggested that a decrease in aerobic capacity, through either an increase in routine metabolic rate or a decrease in maximal metabolic rate, limits thermal tolerance (Pörtner 2002). Support for this hypothesis has come from experiments on water-breathing ectotherms which show that as organisms approach their upper and lower thermal limits, they transition from aerobic to anaerobic respiration, which can eventually lead to death if the organism is not removed from these temperatures (Pörtner et al. 1998; Pörtner 2001). Testing the hypothesis that oxygen limitation is the major determinant of an organism’s thermal tolerance can be difficult; however, previous studies have shown an effect of dissolved oxygen concentration on thermal tolerance limits. In 1962, Alabaster and Welcomme showed that the survival time at lethal temperatures of two species of fish was correlated with dissolved oxygen levels. At low oxygen concentrations, survival time at high temperature was significantly lower than in water with high oxygen levels. In addition, Weatherley (1970) found that the lethal temperature of goldfish could be raised by up to 2°C when concentrations of dissolved oxygen were raised to 5atm or higher. Weatherly (1970) also found that mean survival times of goldfish to lethal temperatures were, on average, 2-5 times greater in the oxygen rich water as compared to normally aerated water. Both of these studies suggest that it may be an inability of fish to 3  obtain oxygen from the environment at high temperatures that initiates their decline to death at high temperatures.  Combined, these findings have led to the formation of the “oxygen-limited thermal tolerance (OLTT)” hypothesis, which links the effects of temperature to performance, at the level of the organism (Pörtner 2002). The OLTT hypothesis states that at both the upper and lower thermal limits, the performance of an organism is limited by the inability of the organism’s respiratory system to match oxygen supply with mitochondrial demand. As temperatures rise, metabolic rate rises, increasing demand, but oxygen solubility in the water decreases and, eventually, mitochondrial demand exceeds the respiratory and circulatory systems’ ability to supply oxygen (Pörtner and Knust 2007). A reduction in aerobic scope results, as metabolic rate increases, which leads to reduced performance in the organism. A critical temperature is reached, eventually, at which point even the standard metabolic rate cannot be maintained and, thus, systems begin to fail, ultimately leading to death (Pörtner 2002).  Pörtner (2001) postulated that the thermal tolerance of an organism is associated with its metabolic rate because an increased resting metabolic rate suggests a higher resting demand, and thus potentially a lower aerobic scope. Therefore, organisms with a higher oxygen demand should exhibit decreased thermal tolerance. However, he went on to suggest that animals could support high metabolic rates by adjusting mitochondrial densities and functional properties in order to shift their thermal tolerance windows. For example, by altering the lipid bilayer composition of the mitochondria, the types of enzymes/isoforms present, and membrane protein composition, organisms can enhance the efficiency of respiration in the mitochondria and therefore change their metabolic demand for oxygen in order to match the supply by the environment (Pörtner 2001). Changing the number of mitochondria in a tissue would also alter 4  the organism’s capacity for aerobic respiration where a decrease in mitochondrial density would lead to an increase in thermal tolerance due to a decrease in oxygen demand. Therefore, mitochondria function or densities could play a major role in setting the thermal tolerance of an organism.  In addition to decreasing thermal tolerance, increased metabolic rate would also be expected to reduce hypoxia tolerance. Most vertebrates are oxy-regulators, meaning they maintain a roughly constant level of “resting” oxygen consumption across a wide range of environmental partial pressures of oxygen, Po2’s (Hughes 1973; McNab 2002). Animals can exhibit many adaptations to supply their metabolic demand under hypoxic conditions, such as increasing ventilation, increasing haemoglobin-oxygen affinity, or utilizing anaerobic pathways. However, at a certain point, termed the critical oxygen tension (Pcrit), the decrease in environmental O2 is too great and the animal must decrease its oxygen consumption (Hughes 1973; McNab 2002; Mandic et al. 2009). Pcrit is often used as a proxy for hypoxia tolerance of species such that animals with high Pcrit’s will be restricted to well-oxygenated environments (Chapman et al. 2002; McNab 2002).  One would predict that organisms with a high metabolic rate, all else being equal, would have a higher Pcrit than organisms with a lower metabolic rate. Genetic influences on phenotype  Variation in physiological phenotypes such as metabolic rate, thermal tolerance, and critical oxygen tension must be related to underlying genetic variation in order to respond to natural selection. One of the classic examples of genetic variation in natural populations leading to differences in physiological phenotype is in phosophoglucose isomerase (PGI) in Colias butterflies.  PGI is an enzyme that converts fructose-6-phosphate to glucose-6-phosphate which links substrates from glycolysis to other pathways, such as gluconeogenesis. Since butterflies use 5  sugar to fuel their flight, Watt (1977) hypothesized that selection for optimal flight performance in response to environmental temperature would, consequently, target PGI. Of the multiple allelic variants found in C. eurytheme, one was found to be optimized for thermal stability and one for substrate-binding affinity (Watt 1977). Watt (1977, 1983) also found that heterozygous butterflies with heterodimeric enzymes, often functioned better than homozygotes across a wide range of environmental temperatures. These findings led to the prediction that heterozygous Colias butterflies should have metabolic pathways that function across a wider range of temperatures and, thus, should be able to fly across a wider range of temperatures (Watt 1983; Watt et al. 1983). This hypothesis was supported in subsequent work on Colias heterozygotes and led to the prediction that these differences in performance may affect fitness of the butterflies (Watt 1983; Watt et al. 1983). The apparent selection on the PGI locus in Colias butterflies is a good example of genetic variation that leads to biochemical differences and, therefore, differences in physiological performance and fitness of the organism.  In the past, as can be seen in the example above, work examining the effects of genetic variation on physiological phenotype has used a candidate gene approach in which biochemical measurements are correlated to genetic variation at carefully selected candidate genes. These biochemical differences are then correlated to phenotypes at the level of the cell or organism (Dalziel et al. 2009). Using a candidate gene approach, however, has its limitations. For example, by screening for candidate genes, loci from well described biochemical pathways may be chosen more often, biasing against lesser-known loci which could also play a role in shaping phenotype (Dalziel et al. 2009).  As a result, Dalziel and colleagues (2009) suggest that background genetic variation must be controlled in order to maximize the ability of an experiment to infer a causal relationship between genotype and phenotype. One way of 6  achieving this is through the use of natural hybrid zones. Individuals from the center of the hybrid zone could be segregating multiple alleles of a candidate gene within a homogenous genetic background, allowing a causal link to be shown between genetic variation at a gene of interest and phenotype (Dalziel et al. 2009). One way to make the link between genotype and phenotype in a hybrid zone is to generate a hybrid index, an estimate of the proportion of alleles in a hybrid individual that came from one of the parental populations. For example, Mousseau and Howard (1998) used species-specific alleles at diagnostic allozyme markers to generate a hybrid index score for crickets from six different populations along a hybrid zone. They found that all measured calling song traits were strongly correlated with hybrid index score and that this correlation was maintained in crickets raised in a common environment. While all of the above examples have involved correlating nuclear genotype with phenotype, variation in the mitochondrial genome may also be correlated with variation in phenotype. Although the mitochondrial genome does not code for nearly as many genes as the nuclear genome, mutation rates in mitochondrial DNA (mtDNA) are approximately 5-10 times faster than in nuclear DNA (Brown et al. 1979; Mortiz et al. 1987; Reichert and Neupert 2004; Burton et al. 2006; Lane 2009). This increased mutation rate means that variants that are better suited to a particular environment will arise more frequently. In addition, mtDNA genes play crucial roles in the cell: proteins encoded for by mtDNA play an important role in cellular energy metabolism, as well as some genes are involved in mitochondrial protein synthesis (Burton et al. 2006). The rapid mutation rate of mtDNA and the role mtDNA plays in cellular metabolism suggests that the mitochondrial genome of an individual may be correlated with physiological phenotypes, such as whole-organism metabolic rate.  7  Interactions between mitochondrial and nuclear genotypes  Mitochondrial and nuclear genes can affect physiological phenotype separately or in combination. Normally, selection works to ensure the two genomes function properly together, such that changes in one genome should select for changes in the other (Blier et al. 2001; Burton et al. 2006). Over time, this process can lead to differentiation between species in both the mitochondrial genome and in nuclear-encoded mitochondrial genes. Physiological deficiencies could occur in hybrid individuals whose nuclear background corresponds to one parental type, but whose mitochondria have been inherited from the other parental type. These cytonuclear incompatibilities could lead to a breakdown (or a reduction in fitness) of hybrid individuals (Edmands and Burton 1999; Burton et al. 2006; Ellison and Burton 2006; Ellison and Burton 2008; Niehuis et al. 2008). While it may be possible to see adverse effects of incompatibilities from just a few amino acid substitutions, high levels of mtDNA sequence divergence between populations will increase the likelihood of these incompatibilities occurring (Burton et al. 2006). The consequences of cytonuclear incompatibility could be manifest in the mitochondrial electron transport chain (ETC), which is responsible for energy metabolism in a cell. The mitochondrial genome of most animals includes 13 protein coding genes that are all critical subunits of the ETC enzyme complexes (Blier et al. 2001; Burton et al. 2006). However, estimates have indicated that there are over 700 proteins in the mitochondrion of metazoans (Reichert and Neupert 2004) which are coded for by the nuclear genome. As a result, there must be functional interactions between nuclear-encoded and mitochondrial-encoded subunits within the ETC which are crucial to proper functioning of the ETC and, consequently, cellular metabolism. 8  The negative effects of cytonuclear disequilibrium on the fitness of hybrids have been seen in a variety of taxa from yeast to humans (Zeyl et al. 2005; King and Attardi 1989). For example, Complex IV (COX) of the ETC was found to have lower activity levels in hybrid frog embryos with the nuclear genome of one parent species, but the mitochondrial genome of the other parent species (Liepins and Hennen 1977). Embryos with a haploid set of chromosomes from the same species as the maternally-inherited mitochondrial genome showed restored COX activity (Liepins and Hennen 1977). A decrease in COX activity was also found in hybrid individuals from crosses between divergent populations of the intertidal copepod, Tigriopus californicus. Edmands and Burton (1999) showed that when introgressing the nuclear genes of one population with the mitochondrial genes of another population, there was a decline in COX activity which was not present in the control backcrosses. Additional work by Burton and colleagues has also found that F2 hybrids of T. californicus showed slower development rates (Burton 1990) as well as reduced fecundity and viability (Edmands 1999) as a result of cytonuclear incompatibilities. Interestingly, the magnitude of the detriment to fitness seemed to show some dependence on geography and/or genetics with more divergent population pairs leading to greater hybrid breakdown, although the relationship was not consistently strong (Burton et al. 2006). While the exact proteins responsible for hybrid breakdown are difficult to determine, examining the effects of cytonuclear incompatibilities on enzyme activities, or even whole animal physiological measurements, such as metabolic rate, in natural populations offers some advantages. Blier and colleagues (2001) argued that although direct manipulation of cells in culture, or producing lab-reared hybrids, might offer greater control and repeatability, natural hybrid populations may be able to provide insight on the effects of environment on hybrid 9  breakdown. In addition, whole-animal measures of physiology may be more relevant to the fitness of an organism in nature than the interactions occurring at a microscopic level. Since processes such as mitochondrial metabolism are often thought to evolve in response to environmental factors, such as temperature (Blier et al. 2001), physiological studies of hybrid individuals from heterogenous environments, such as an intertidal zone, may provide better insight to the consequences of cytonuclear incompatibilities to the animal as a whole. Little work has been done, however, on the effects of cytonuclear incompatibilities on whole-animal physiology: one of the goals of this thesis is to illuminate this area of research. Killifish as an experimental model  To address my thesis objectives, I have used sub-species of Fundulus heteroclitus, the common killifish, which inhabits marshes and estuaries along the east coast of North America from Newfoundland to northern Florida (Hardy 1978; Morin and Able 1983). As a result of this wide geographic distribution, killifish are exposed to a range of environmental temperatures and salinities and demonstrate a remarkable tolerance to environmental conditions. It may be because of this range in environmental conditions that killifish are a very hardy species, even in captivity (Atz 1986). The abundance of killifish in the wild and the ability to successfully breed and maintain large populations in the laboratory have made killifish an ideal research organism in many laboratories across North America (Atz 1986). Killifish are also ideal model species because of their compact size and relatively short life span: adults range in total length from 50 to 100 mm and have an average life span of 4 years (Abraham 1985). Killifish are sexually dimorphic, with males exhibiting banding patterns and exaggerated colouration during the breeding season (Hardy 1978; Abraham 1985). Fundulus heteroclitus utilize a semilunar reproductive strategy in which females deposit their eggs during a spring tide to hatch during the 10  next high spring tide approximately two weeks later (DiMichele and Westerman 1997). Eggs are incubated aerially and are triggered to hatch when immersed in water (Abraham 1985). Killifish habitat  Fundulus heteroclitus is one of the most abundant resident fish species in estuaries along the East Coast of North America (Taylor 1999; Sweeney et al. 1998). Daily movements of killifish between tide pools and marshes vary with the tides as well as seasonally. For example, Smith and Able (1994) found that killifish from Southern New Jersey were more commonly found in shallow, salt-marsh tide pools during the winter months instead of the intertidal and subtidal creeks which they were more commonly found in during the summer. While killifish do move between marshes, they also exhibit high site fidelity: various mark-recapture experiments have determined their home range to be anywhere from 18 m to 2 km (Fritz et al. 1975; Lotrich 1975; Skinner et al.  2005; Sweeney et al. 1998). Evidence from these studies suggests that the small home ranges exhibited by F. heteroclitus may be correlated with the amount of habitat available for reproduction and predator avoidance as well as the availability of food (Lotrich 1975; Skinner et al. 2005).  Despite the fact that an individual fish is restricted to a small geographic range, the species as a whole experiences a wide range of environmental conditions. For example, the natural distribution of F. heteroclitus encompasses one of the largest thermal gradients in the world, with mean surface water temperatures changing at a rate of 1°C per degree of latitude (Powers and Place 1978). At the extreme ends of the killifish’s distribution the mean monthly temperature differences are approximately 13°C with an overall temperature range of -1.4 to 31.6°C (Powers et al. 1991). In addition to geographical and seasonal temperature gradients, 11  there are also daily fluctuations in temperature. For the majority of the year, daily temperature can vary between 5°C below and above the mean daily temperature. Spatial variation in temperature is also a factor due to the fact that the coastal waters, estuaries, and tide pools could all differ in thermal regimes- even within a small geographical area. However, temperature is not the only variable in the killifish environment- dissolved oxygen, pH, and salinity also vary daily and seasonally in intertidal marine habitats (Truchot and Duhamel-Jouve, 1980). This diversity in native habitat conditions makes killifish an excellent model for studying environmental adaptation and also allows experimenters the opportunity to investigate the effects of sample location on phenotypic traits. Killifish genetic variation Controversy has surrounded the taxonomic status of Fundulus heteroclitus for nearly a century. Previous studies have demonstrated substantial variation in morphological, molecular, genetic, and physiological traits within the species all of which show significant directional change with temperature and latitudinal gradients (Morin and Able 1983; Mitton and Koehn 1976; Powers et al. 1986; Powers et al. 1993; Powers and Schulte 1998; Schulte 2001; Scott et al. 2004). This pattern of variation has led to the identification of two distinct subspecies: the northern form, Fundulus heteroclitus macrolepidotus, which ranges from Newfoundland, Canada to northern New Jersey, USA, and the southern form, Fundulus heteroclitus heteroclitus, which is distributed from Virginia, USA to the North-eastern coast of Florida, USA (Morin and Able 1983; Able and Felley 1986). In addition to morphological and microsatellite data, analysis of mitochondrial DNA (mtDNA) has shown that the Northern and Southern haplotypes diverged approximately 1 million years ago (Bernardi et al. 1993; Gonzalez-Villasenor and Powers 1990), again providing reassurance for the designation of a Northern and Southern form as distinct 12  subspecies. Interestingly, the clines in mitochondrial and nuclear genotype allele frequencies all transition from one predominant form to another around 40-41°N latitude in Northern New Jersey, providing even more compelling evidence for the separation of F. heteroclitus into two subspecies (Powers and Place 1978; Bernardi et al. 1993; Adams et al. 2006). While both mitochondrial and nuclear DNA patterns within the species show a clear divergence between the Northern and Southern forms of killifish, they also indicate that the contact zone in Northern New Jersey exhibits high rates of recombination and the greatest amount of genetic diversity (Powers and Place 1978; Gonzalez-Villasenor and Powers, 1990; Bernardi et al. 1993; Adams et al. 2006; Duvernell et al. 2008). The abrupt break in genetic variation between the subspecies, near the Hudson River in Northern New Jersey, suggests that interbreeding between the northern and southern form occurs in this contact zone leading to hybrid individuals (Bernardi et al. 1993; Gonzalez-Villasenor and Powers 1990; Adams et al. 2006). The existence of hybrid individuals was also suggested by Morin and Able (1983) when they noticed that egg characteristics, such as the number of oil droplets, were intermediate between Northern and Southern forms in fish from Northern New Jersey, despite the fact that egg morphology is usually very well conserved within a species. Both the genetic and morphological data provide a compelling argument for the existence of hybrid individuals between Northern and Southern killifish. The substantial genetic variation in both nuclear and mitochondrial DNA loci between the two subspecies suggests that both spatially variable selection and secondary intergradation have played a role in the present-day species distributions (Powers and Schulte 1998). While the exact mechanism of divergence in still unclear, researchers believe that the current geographic distribution and genetic variation between the subspecies is a result of one of two hypotheses: 13  the northern expansion hypothesis and the northern refugium hypothesis (Powers et al. 1986; Powers and Schulte 1998; Ropson et al. 1990; Wares 2002; Adams et al. 2006). The northern expansion hypothesis claims that killifish were absent from the northern half of their current geographic range during the Pleistocene due to extensive glaciation (Powers et al. 1986; Ropson et al. 1990; Powers and Schulte 1998; Adams et al. 2006). Upon glacial retreat, the more cold- tolerant individuals made a step-wise expansion through the northern part of the range as new habitat became available (Powers et al. 1986; Powers and Schulte 1998; Ropson at al. 1990; Adams et al. 2006). The northern refugium hypothesis, proposed by Wares in 2002, suggested that the Northern populations of killifish were isolated during the Pleistocene in a refugium in western Nova Scotia near the Bay of Fundy (Wares 2002; Adams et al. 2006). As glaciers melted, these geographically isolated Northern forms moved southwards and made secondary contact with Southern populations (Wares 2002; Adams et al. 2006). However, recent microsatellite evidence seems to reject both of these hypotheses (Adams et al. 2006). Instead, it has been suggested that killifish actually inhabited a range similar to the present-day latitudinal distribution, but were located further out in the ocean (Adams et al. 2006). As glaciers melted, sea-level increased substantially, decreasing the amount of shoreline distance between Nova Scotia and New Jersey which may have created disequilibrium conditions as the Northern and Southern forms colonized new habitats (Adams et al. 2006). All of the proposed mechanisms for divergence lead to the formation of two subspecies as a result of either latitudinal distance, or environmental conditions, such as temperature. As a result, genetic variation within the species is strongly correlated with latitude, with the Northern subspecies having much lower genetic diversity than the Southern subspecies (Ropson et al. 1990; Adams et al. 2006; Duvernell et al. 2008). Adams et al. (2006) detected almost no private 14  Northern microsatellite alleles, whereas none of the common microsatellite alleles from the Southern subspecies extended into the Northern subspecies reinforcing the subdivision of Fundulus heteroclitus into two subspecies. These results were concordant with clines in allozyme frequency found 30 years earlier (Powers and Place 1978). Powers and Place (1978) found clines in malate dehydrogenase (MDH), phosphoglucomutase (PGM), and lactate dehydrogenase (LDH-B) allozyme frequencies. The cline in lactate dehydrogenase has been the focus of the majority of studies examining allozyme frequency clines in Fundulus. LDH-B is an enzyme that catalyzes the interconversion of pyruvate (used in aerobic glycolysis) and lactate (the end product of anaerobic glycolysis). Killifish have two codominant alleles of Ldh-B4 which are virtually fixed between the two extremes of the range: LDH-B N , which is predominant in the Northern subspecies, and LDH-B S , which is predominant in the Southern subspecies (Powers and Place 1978). Upon examination of their catalytic efficiencies, Place and Powers (1979 and 1984) found that the Northern enzyme, LDH-B NN , had a greater efficiency at lower temperatures, which is consistent with the idea that the Northern subspecies has undergone local adaptation to low temperatures. The Southern genotype, LDH-B SS , however, did not show an increased efficiency at warm temperatures. Thus, there appears to be a correlation between genotypic variation and local adaptation to environment, at least in the case of LDH-B allozymes. Because of the genetic variation between the subspecies and their wide latitudinal and thermal range along the East Coast, killifish have often been used to examine the role of temperature as a selective factor. Multidisciplinary approaches including population genetic analysis, phylogenetic independent contrasts with other closely related Fundulus species, and selection experiments have all shown that genetic variation between killifish subspecies, as well 15  as their physiological specializations, are likely to be adaptive responses to temperature or other variables correlated with latitude (reviewed in Powers and Schulte 1998; Schulte 2001). Brown and Chapman (1991) suggested that the clinal distributions in allozymes and morphology are most likely maintained by natural selection, an idea that is supported by the findings of Place and Powers (1979 and 1984). The extent to which differences in biochemical phenotype reflects adaptation to environmental temperature was also examined by Mitton and Koehn (1975) using three populations of killifish: a Northern population, Southern population, and a Northern population living in the heated effluent of a power plant which created water temperatures that were more similar to those experienced by Southern populations than by their native Northern populations. The Northern population living in the heated effluent had isozyme patterns (allele frequencies) which more closely resembled the Southern populations (Mitton and Koehn 1975). This suggests that protein polymorphisms in the killifish can be an adaptive genetic response to temperature and, therefore, environmental influences on phenotype may actually be reflected in genotype. Killifish physiological variation  In addition to latitudinal clines in genetic variation, killifish also exhibit variations in physiology and biochemistry along temperature and geographic gradients. Previous studies have shown differences between Northern and Southern populations of F. heteroclitus in mitochondrial properties and enzyme activities, whole animal metabolic rate, critical thermal maximum, and critical oxygen tension as well as other physiological parameters (Fangue et al. 2006; Richards et al. 2008; Fangue et al. 2009a).  The extensive geographic range as well as genetic differentiation between the subspecies makes the killifish an ideal model system for studying the effects of environment and genotype on physiology of an organism. 16   Much of the previous physiological work on killifish has focused on mitochondrial properties of the subspecies as well as metabolic rates. Work done by Fangue and colleagues (2009a) showed that mitochondrial respiration rates were higher in the putatively cold-adapted Northern fish at lower acclimation temperatures. The opposite was true, however, at warmer test temperatures.  The same study also found that when Northern and Southern killifish were acclimated to 5°, 15°, and 25°C and tested at these temperatures, Northern killifish consistently had a significantly higher whole-animal mass-specific metabolic rate than Southern killifish. In addition, Northern killifish were found to have significantly higher maximum rates of mitochondrial oxygen consumption than Southern killifish at cooler acclimation temperatures, suggesting that the Northern subspecies may have a greater capacity for mitochondrial function (Fangue et al. 2009a). Although all of these differences seem to be more apparent at cold temperatures, there are clearly differences in mitochondrial function between the subspecies which likely play a role in the differences seen at the whole-animal level.  Previous studies have also examined the differences in thermal tolerance, as well as thermal preference, of killifish subspecies. Fangue et al. (2006) examined the thermal tolerance of Northern and Southern killifish across a wide range of acclimation temperatures (2-34°C) and found that critical thermal maximum was approximately 1.5°C higher in the Southern fish than the Northern fish across all acclimation temperatures. The critical thermal minimum was also approximately 1.5°C higher in Southern fish than Northern fish, but only at acclimation temperatures above 22°C. This decrease in the upper thermal tolerance limit seen in the Northern subspecies, appears to be inversely correlated with the increased metabolic rate seen in the group compared to the Southern subspecies (Fangue et al 2009a). These findings provide support for 17  the OLTT hypothesis and suggest that the limits to Fundulus’s thermal niche may be caused by limits in their scope for aerobic activity. In addition, Fangue et al. (2006) examined whether these differences in thermal tolerance were reflected by differences in heat shock protein (hsp) sequence or expression. While they only found 1 conservative amino acid substitution in hsc70, after looking at 5 different hsp genes, they did see a difference in expression patterns between the populations. These findings support the idea that temperature, or some other factor correlated with latitude, has led to an adaptive response by killifish which is reflected by the physiological and genetic variation between the subspecies (Fangue et al. 2006). Temperature preference has also been measured in each subspecies. Temperature preference of Northern and Southern subspecies acclimated to 5°, 15°, and 25°C was determined to be significantly higher in Northern fish than in Southern fish (Fangue et al. 2009b). When high-latitude organisms prefer higher temperatures, this pattern is called countergradient variation. This pattern predicts that organisms at higher latitudes will prefer higher temperatures in order to increase performance, growth and fecundity and take advantage of shorter growing and reproductive seasons. This could result in an increase in fitness of the Northern subspecies compared to lower latitude equivalents when living in the same habitat (Fangue et al. 2009b). Evidence for countergradient variation in killifish suggests that there may even more physiological differences between the subspecies which have yet to be measured. While past research has not examined differences in hypoxia tolerance between the subspecies, the oxygen carrying capacity of the blood in both subspecies has been examined. Mied and Powers (1978) showed that the major allosteric modulator of haemoglobin-oxygen binding in fish is ATP, which decreases haemoglobin-oxygen affinity. Subsequent work by 18  Powers and colleagues (1979) found that the Southern individuals who were homozygous for the LDH-B S  allele had a lower ATP/Hb ratio in the red blood cell and, consequently, a higher haemoglobin-oxygen affinity than Northern individuals who were homozygous for the LDH-B N  allele. Heterozygous fish were found to have intermediate intracellular ATP concentrations (Powers et al. 1979). These findings led to the prediction that fish with the LDH-B SS  phenotype should be able to extract more oxygen from a hypoxic environment and, thus, will have a selective advantage over other phenotypes (Powers et al. 1979). This means that Southern individuals, who are homozygous for the LDH-B S  allele should also exhibit a lower critical oxygen tension (Pcrit). Although Pcrit values have never been compared between the subspecies, Northern killifish were found to have a Pcrit of 63.9 ± 5.9 torr at 20°C: a number that was surprisingly high for a species that is well-known for its hypoxia tolerance (Richards et al. 2008). In addition, Mandic et al (2009) found that 75% of the variation in Pcrit among species of sculpin was explained by variation in routine metabolic rate, mass-specific gill surface area and whole blood haemoglobin-oxygen binding affinity (P50). Combined, these findings, along with those of Fangue (2009a) and Powers (1979), indicate that Northern and Southern subspecies of killifish may have different critical oxygen tensions. Determining the relative Pcrit values of both subspecies could help support Powers claims that Southern fish, with the LDH-B SS  phenotype, are better adapted for hypoxia. The LDH-B allozyme differences between the subspecies are correlated with many other physiological differences as well, many of which have been identified in embryos. Killifish utilize a semilunar reproductive strategy, where eggs are deposited during a high tide and hatch during the next high tide approximately two weeks later after developing in air (DiMichele et al. 1986). As a result of temperature differences between the extreme ends of the killifish’s range, 19  one would expect that differences in embryonic development could be seen between Northern and Southern killifish. As predicted, differences were found in development rate of embryos, where individuals homozygous for the LDH-B N  allele were shown to have a decreased embryonic development rate (DiMichele and Powers, 1982a). Homozygous LDH-B N  individuals also showed increased hatching time, better swimming endurance, lower metabolic rates, decreased lactate metabolism and decreased glucose production (DiMichele and Powers, 1982a,b; DiMichele and Powers 1984; DiMichele et al. 1986; Paynter et al. 1991). To ensure the observed physiological differences were a result of differences in LDH-B allozyme type, DiMichele et al. (1991) injected LDH-B SS  enzyme into an LDH-B NN  egg and found that metabolic rate and glucose flux were determined by the injected enzyme. Previous work done on the LDH-B locus, at least, gives compelling evidence of the role of genotype on phenotype.  The physiological differences between the Northern and Southern subspecies of killifish lead to interesting questions about the physiological phenotypes of fish from the contact zone. No studies, at present, have quantified the metabolic rate, thermal tolerance or critical oxygen tension of adult hybrid individuals. The killifish model system affords the unique opportunity to study the effects of hybridization on physiological phenotype in naturally occurring hybrid individuals. The use of hybrid individuals is especially intriguing given the correlation between LDH-B and physiological variation found in previous work. This study will determine whether genotype of an individual killifish, at both neutral nuclear markers as well as mitochondrial genotype, is correlated to its physiological phenotype. If a correlation is found between genotype and phenotype, it would suggest that we could predict an individual’s fitness and performance in the lab merely through its genetic background.  20  Evidence for cytonuclear incompatibilities in killifish While differences in mitochondrial function may be related to whole-organism metabolic rate, no studies have examined mitochondrial properties of fish from the contact zone. It is unclear how killifish from the contact zone with mismatched mitochondrial and nuclear genotypes, with respect to components of the ETC, would be affected by cytonuclear incompatibilities on a whole-animal level. However, Blier et al. (2001) suggest that cytonuclear incompatibilities might be more common in cases where an organism’s distribution includes a wide range of thermal habitats and where enzyme variants may be selected for as a result of having increased efficiency at a given temperature, as has been observed in the different LDH-B alleles of killifish (Place and Powers 1978; Blier et al. 2001). Burton and colleagues (2006) suggested that a good candidate genetic system for examining the effects of cytonuclear disequilibrium could be involved in central metabolism since disruptions of biochemical pathways may result in reduced energy production. They proposed that a great differentiation in mtDNA genes may disproportionately affect fitness and physiology of hybrids because of their functional role in cellular metabolism (Burton et al. 2006). The divergence of killifish mtDNA, along with the observations made by Fangue and colleagues on killifish mitochondrial function and metabolic rate, makes killifish an ideal model system for examining not only the effects of nuclear and mitochondrial genotypes on physiology, but also the effects of incompatibilities between the two on physiology. Thesis objectives  Previous work on killifish supports the prediction of the oxygen-limited thermal tolerance hypothesis that there is a negative correlation between metabolic rate and thermal 21  tolerance, but little is known about either the physiological variation or the genetic variation of fish from the contact zone. My work examines this correlation in hybrid individuals. By examining the metabolic rate and thermal tolerance of fish from a small geographic region with a range of nuclear and mitochondrial genotypes, I will be able to more clearly determine whether or not this correlation represents a functional relationship and also whether mitochondrial genotype, nuclear genotype, or mismatches between the mitochondrial and nuclear genome affect the physiological phenotypes.  The purpose of my thesis was to determine if metabolic rate and thermal tolerance were correlated in Fundulus heteroclitus from a hybrid zone between the two Fundulus subspecies and to investigate whether or not genotype has an effect on the thermal tolerance, metabolic rate and critical oxygen tension. Specifically, I had four main research questions: 1) does mitochondrial genotype or collection location affect thermal tolerance, metabolic rate, or Pcrit, 2) does nuclear genotype affect thermal tolerance, metabolic rate, or Pcrit, 3) does a mismatch between nuclear genotype and mitochondrial genotype affect thermal tolerance or metabolic rate and 4) are thermal tolerance and metabolic rate inversely correlated in fish from the contact zone? 22  MATERIALS AND METHODS- Animals  Adult killifish of the northern subspecies were collected from Hampton, NH, USA (NH; 42° 54’ 46” N) by Aquatic Research Organisms (Inc.). Adult killifish of the southern subspecies were collected from Brunswick GA, USA (GA; 31° 7’ 31” N) and Morehead City, NC (NC; 34° 43’ 44’’ N) by L. Glass (North Carolina State University).  All Northern fish as well as the Southern fish from Georgia were collected in spring of 2008, while Southern fish from North Carolina were collected in summer of 2008. Adult killifish from the region of overlap between the two subspecies were collected from an estuary at the mouth of the Navesink River (Red Bank, NJ; 40° 24’ 0” N) and from a tributary of the Metedeconk River (Beaver Dam Creek, Point Pleasant, NJ; 40° 03’ 0” N) in May of 2008 using un-baited Gee’s G40-type minnow traps. All collected fish were sent to Aquatic Research Organisms in Hampton, NH where they were held and acclimated to laboratory conditions before being transported to the University of British Columbia (UBC). At UBC, all fish were acclimated to the laboratory in aquaria containing artificial seawater at a salinity of approximately 20‰, achieved using KENT Sea Salt (KENT Marine, Franklin, WI, USA). Prior to experimentation, fish were acclimated to 20°C for at least three weeks with a 12 light: 12 dark photoperiod. Fish were fed to satiation once daily using TetraMin ®  fish flakes. Water quality was monitored and water changes were performed as necessary. Each fish used in experimentation was marked subcutaneously using Visible Implant Elastomer (Northwest Marine Technology, Inc., Shaw Island, Wash., USA) and allowed to recover for at least one week prior to the start of experiments. The combination of tank assignment and elastomer colour and location created a tagging scheme that allowed each individual to be identified. All experimental procedures were approved by The University of 23  British Columbia Animal Care Committee on behalf of the Canadian Council on Animal Care and were conducted under approved protocol #A07-0288. Whole-animal respirometry  Routine metabolic rate (Mo2) and critical oxygen tension (Pcrit) were determined using closed respirometry. The apparatus consisted of four glass jars of approximately 220 ml capacity that were sealed with two-holed rubber stoppers. Each vessel contained a magnetic stir bar and was placed over an individual stir plate in order to ensure proper mixing of water inside the respirometer. A stopcock with a tube from a well-aerated header bucket, containing water at a salinity of 20‰ and a temperature of 20°C, was placed over one hole while a fibre-optic oxygen probe (FOXY-R, Ocean Optics Ltd., Dunedin, FL, USA) was fitted into the second hole. Prior to measuring Mo2 and Pcrit, one fish was placed in each respirometer overnight (16-20 hours) to habituate to the apparatus and reduce the effect of handling stress. During this time, the respirometer remained open, allowing air-saturated water to flow into the respirometer. At the start of a trial, the stop cock was closed and the probe was placed in the respirometer, creating an air-tight seal with the rubber stopper. The temperature of water inside the chambers was regulated using a 20°C water bath for the duration of the measurement. The decline in dissolved oxygen was recorded continuously.  Each fish remained in the closed respirometer until it reached a partial pressure of oxygen of approximately 10-15 torr or until the fish began to show signs of distress. Trials took between 45 and 120 minutes and fish were generally quiescent for this duration. Any fish that struggled during the trial, or had an oxygen trace that was too variable to calculate a stable routine metabolic rate from, was removed from the data set. At the conclusion of each trial, the fish was removed from the chamber weighed, measured, and placed back in its acclimation tank. 24  Mass-specific routine metabolic rate (Mo2) was calculated following Henriksson et al. (2008). The differential solubility of oxygen in water of varying salinities was corrected for using the solubility coefficients, αO2, at a constant temperature of 20°C (Boutilier et al. 1984). The slope of the oxygen trace (water [O2] over time) was calculated over sequential 5 minute intervals and corrected for fish wet weight and respirometer volume to give Mo2 (µmol/g/h). Routine Mo2 was calculated by using the slope of the shallowest portion of the dissolved oxygen trace at partial pressures of oxygen (Po2) greater than 60 torr where that slope was maintained for at least a 7 minute interval. Calculated Mo2’s for each 5 minute interval were then plotted against the mean water Po2 for that interval and Pcrit was defined as the inflection point where Mo2 transitioned from being independent of environmental Po2 to being dependent on environmental Po2 (Pörtner and Grieshaber 1993). Pcrit was calculated by determining the equation for the linear regression through the calculated Mo2’s that were more than 12% below the calculated routine metabolic rate.  We used this approach because pilot studies on Mo2 repeatability showed that we could accurately determine the routine Mo2 of an individual to within ± 12% (Appendix, Figure A1 and A2). Pcrit was determined by calculating the Po2 at the point where the linear regression reached the routine metabolic rate. Thermal tolerance  We used Critical Thermal Methodology (CTM) to determine the maximal thermal tolerance, or, CTMaximum (CTMax), of fish from the northern, southern, and intermediate populations. The CTMax is an estimate of the temperature at which a fish is no longer able to escape from potentially deadly conditions (Becker and Genoway 1979; Beitinger et al. 2000; Cox 1974). A CTMax trial is conducted by gradually increasing water temperature and then recording the temperature at a specific end point. Here we used heating rates of 0.3°C/min in all 25  trials and used Loss of Equilibrium (LOE) as an endpoint. LOE is the point at which a fish can no longer remain dorso-ventrally upright. CTMax was calculated by taking the arithmetic mean of the LOE temperatures for each population (Cox 1974; Beitinger et al. 2000). At the conclusion of each trial, each fish was weighed, measured and a small caudal fin clip was taken for use in genetic analysis. The fish were returned to their acclimation tanks for recovery. Fish that died between metabolic rate trials and thermal tolerance trials were excluded from analysis. No mortality was observed within two weeks of the conclusion of these trials. The CTM apparatus consisted of a plastic rectangular water bath with a Plexiglass lid containing 12 holes into which individual 1 l plastic chambers with Plexiglass lids could be placed. The water bath was heated by adjusting the rate of water flow from a header tank. The header tank was filled with hot tap water (approximately 50°C) which was kept hot using a Lauda RM6 benchtop unit. A pump was used to feed water from the header into the water bath and a plastic valve was used to control the rate of flow from the header tank. Excess water was allowed to flow out of the bath and into a floor-level drain via a hole in the plastic water bath that was just below the top of the individual beakers. A large airstone was used to prevent thermal stratification in the header tank. Water in the water bath was kept well-mixed by two pumps as well as by a large air stone placed in front of the line of flow of both pumps. Each individual beaker was filled with 20‰ water and was aerated with an individual airstone to ensure complete oxygen saturation and prevent thermal stratification with increasing temperatures during the trials. Heating rates in each column of beakers were monitored using digital temperature controllers (Fisher Scientific, Ottawa, ON, Canada) and beaker temperatures were monitored using a Fisherbrand ®  NIST certified mercury thermometer (Fisher Scientific, Ottawa, ON, Canada). 26  Microsatellite analysis and hybrid index determination  Caudal fin clips were taken from each individual and genomic DNA was extracted using DNeasy Blood and Tissue Kit (Qiagen). Eight trinucleotide (ATG) microsatellite loci, identified by Adams et al. (2005) were used to genotype all individuals (Red Bank n=40; Metedeconk n=36) using polymerase chain reaction (PCR). Three multiplex 12.5 µl PCR reactions were performed: Reaction 1 included the FhATG-17, FhATG-18 and FhATG-20 primers; Reaction 2 included the FhATG-2, FhATG-4 and the FhATG-B103 primers; and Reaction 3 consisted of the FhATG-B101 and the FhATG-B128 primers. One primer from each pair was fluorescently labelled (6-FAM, VIC, NED, or PET). Table 1 shows the PCR master mix cocktails for each reaction.  Table 1. Master mix ingredients in volumes and concentrations for microsatellite multiplex PCR Reactions 1, 2, and 3.  Amount in Master Mix PCR ingredient: Reaction 1 Reaction 2 Reaction 3 MgCl2 0.8 mM 0.8 mM 0.8 mM dNTP 0.4 mM 0.4 mM 0.4 mM primers 0.4 mM 0.4 mM (0.6 mM for ATG103 Forward and Reverse primers only) 0.4 mM TAQ 0.5 units 1 unit 0.5 units 10x buffer 1.25 µl 1.25 µl 1.25 µl dd H2O 5.9 µl 5.3 µl 6.9 µl genomic DNA 1 µl 1 µl 1 µl Total volume: 12.5 µl 12.5 µl 12.5 µl  27   All PCRs were carried out in a PTC-200 MJ Research thermocycler using Taq DNA polymerase (MBI Fermentas Inc., Burlington, ON, Canada). The amplification profile for Reaction 1 consisted of 5 cycles of 30 s at 94°C, 45 s at 54°C, and 1 min at 72°C followed by 30 cycles of 30 s at 94°C, 45 s at 59°C, and 1 min at 72°C. Amplification profiles for Reaction 2 and 3 consisted of 35 cycles of 30 s at 94°C, 45 s at 58°C, and 1 min at 72°C. PCR products were sent to the Molecular Genetics Core Facility at Children’s Hospital in Boston for electrophoresis on an Applied Biosystems 3730 DNA Analyzer (Foster City, CA, USA). Microsatellite alleles were scored by hand using Peak Scanner version 1.0 (Applied Biosystems). Tests for the presence or absence and frequency of null alleles as well as typographic errors were performed using MicroChecker version 2.2.3 (Van Oosterhout et al. 2004). Observed (HO) and expected (HE) heterozygosities as well as tests for linkage disequilibrium among loci were determined using GENEPOP version 3.4 (Raymond and Rousset 1995).  A hybrid index was calculated using Introgress version 1.2 (Gompert and Buerkle 2010). Introgress is written in the R language and was used with R version 2.9.1 (CRAN; http://cran.r- project.org/). It predicts the probability of a given genotype (hybrid index) using multinomial regression to estimate change in frequency of marker genotypes, called genomic clines, based on user designated parental populations (Gompert and Buerkle 2010). Although previous studies have often used Bayesian admixture proportions as an estimate of hybrid index (e.g. Q from Structure; Pritchard at al. 2000), Gompert and Buerkle (2009) suggest that a more accurate hybrid index is obtained when hybrid allele frequencies are estimated based on parents which are defined a priori. Individuals from estuaries on either extreme of the contact zone were used as the parental population.  28  Mitochondrial haplotype determination  A technique utilizing restriction digestion, developed by Jessica McKenzie (unpublished) was used, with permission, to rapidly determine mitochondrial haplotype. Approximately 1100 base pairs (bp) of F. heteroclitus mitochondrial D-loop was amplified via PCR using the forward primer 5’-AGC TCA GCG CCA GAG CGC CGG TCT TGT AAA-3’ and the reverse primer 5’- CGT CGG ATC CCA TCT TCA GTG TTA TGC TT-3’ (Lee et al. 1995). 12.5 µl PCR reactions consisted of 0.4 mM of each primer, 0.2 mM of dNTPs, 0.16 mM of MgCl2, and 1 unit of Taq DNA polymerase (MBI Fermentas Inc., Burlington, ON, Canada) as well as 10x buffer and enough distilled deionized H2O to yield an 11.5 µl reaction. 1 µl of genomic DNA was added to each reaction to yield 12.5 µl total volume per reaction.  The amplification profile consisted of 5 cycles of 30 s at 94°C, 45 s at 60°C, and 1 min at 72°C followed by 34 cycles of 30 s at 94°C, 45 s at 65°C, and 1 min at 72°C, using a PTC-200 MJ Research thermocycler. For the restriction digest reaction, 5 µl of PCR product was digested with ScaI restriction enzyme (MBI Fermentas Inc., Burlington, ON, Canada) for 3 hours at 37°C. This restriction enzyme is designed to cut at a site that is an A nucleotide in the Southerns while leaving the Northerns (whose genome has a G at this loci) intact. The product of the restriction digest was then electrophoresed on a 1.5% agarose gel stained with ethidium bromide and visualized using a UV transluminator (SYNGENE, Frederick, MD, USA). Mitochondrial haplotype was determined by the presence of one band (Northern haplotype) or two bands (Southern haplotype) for each individual. Statistics A linear regression analysis was performed, using Sigma Stat version 3.5 (Systat Software, San Jose, CA, USA), to determine the ability of the logarithm of wet weight to predict 29  Mo2, CTMax, and Pcrit. Sigma Stat was also used to calculate the arithmetic means of Mo2, CTMax, and Pcrit for each group: Northern fish, New Jersey fish, and Southern fish. A One-way Analysis of Variance (ANOVA) was performed, to determine if there were any significant differences between groups. T-tests were performed, also using Sigma Stat, to determine the effects of mitochondrial genotype and sample location on both CTMax and Mo2. Pearson correlations were calculated in Sigma Stat to determine if there was any correlation between nuclear genotype and Mo2, CTMax, and Pcrit. Pearson correlations were also used to determine if there was a correlation between phenotypes, for example, between Mo2 and CTMax. All data were log-transformed, as needed, to fit the assumptions of normality and equal variance. In cases where log-transformed data still did not meet the assumptions of normality and equal variance, the equivalent non-parametric test was used. All data are expressed as means plus or minus the standard errors of the mean (SEM). The alpha value for all tests was set to 0.05. 30  RESULTS- mass as a predictor of physiological phenotype   Individuals were grouped into three categories: Northern fish (n=19), New Jersey fish (putative hybrids, n=63), and Southern fish (n=16) based on their sampling location. Because metabolic rate is known to be influenced by body size (Kleiber 1932), we determined whether there were significant correlations between the logarithm of wet weight and our three physiological phenotypes: metabolic rate (Mo2), thermal tolerance (CTMax), and critical oxygen tension (Pcrit). The logarithm of wet weight of fish was a good predictor of routine metabolic rate (p < 0.001; F=83.399; df=1; Figure 1A), but not CTMax (p=0.298; F=1.098; df=1; Figure 1B) or Pcrit (p=0.625; F=0.240; df=1; Figure 1C). The mean wet weight of Northern fish was 5.697 ± 0.387g; New Jersey fish 5.387±0.278g; and Southern fish 6.279 ± 0.457g. Although mass was not significantly different among populations (ANOVA: p=0.156; F=1.893; df=2) residuals of Mo2 were determined by calculating the predicted metabolic rate from the linear regression equation (Figure 1A) because wet weight was strongly correlated with Mo2. These residuals of Mo2 were used in all statistical calculations of Mo2, unless otherwise stated. 31       Figure 1A. Linear regression between the logarithm of wet weight and the logarithm of routine metabolic rate of all Northern (n=19), New Jersey (n=57) and Southern fish (n=16) (p < 0.001; F=83.399; df=1). y= 0.834x + 1.050 R 2 = 0.475 P < 0.001 32       Figure 1B. Linear regression between the logarithm of wet weight and thermal tolerance (CTMax) of all Northern (n=17), New Jersey (n=54) and Southern fish (n=12) (p=0.298; F=1.098; df=1). Analyzing each population separately also showed no significant correlation between wet weight and CTMax. y= -0.769x+36.831 R 2 =0.0134 p=0.298 33        Figure 1C. Linear regression between the logarithm of wet weight and critical oxygen tension (Pcrit) of all Northern (n=15), New Jersey (n=54) and Southern fish (n=14) (p=0.625;  F=0.240; df=1).  y=4.045x+36.831 R 2 =0.00296 p=0.625 34  Phenotype of New Jersey fish At 20°C, Northern fish, New Jersey fish and Southern fish had residuals of log Mo2 that were not significantly different (p=0.291; F-1.250; df=2; Figure 2A). The mean residual value for Northern fish (n=19) was 0.0362 ± 0.0339 µmol O2/hr; New Jersey fish (n=57) had a mean residual value of 0.0000  ± 0.0179 µmol O2/hr; and Southern fish (n=16) had a mean residual value of -0.0373 ± 0.0270 µmol O2/hr. Mass-specific metabolic rate for each of the three groups is given in Table 2 to facilitate comparison with other studies. Thermal tolerance of these three groups, however, was significantly different with Southern fish (n=12) having the highest CTMax (p<0.001; F=16.223; df=2; Figure 2B) at 37.583 ± 0.198 °C. Southern fish had a significantly different thermal tolerance than both New Jersey (n=54) and Northern fish (n=17) (p<0.001 for both Student-Neuman-Keuls trials). The CTMax of Northern fish (36.218 ± 0.201 °C) and New Jersey fish (36.006 ± 0.124 °C), however, did not differ significantly (p=0.378 by SNK). Critical oxygen tension, Pcrit, like Mo2, showed no significant difference between groups (p=0.262; F=1.364; df=2) at 20°C (Figure 2C). Northern fish (n=15) had a Pcrit of 34.487 ± 1.987 torr; New Jersey fish (n=55) had a Pcrit of 36.803 ± 1.695 torr; and Southern fish (n=14) had a Pcrit of 40.126 ± 2.098 torr. 35   A B C a a a a a b a a a Figure 2. Mean residuals of log Mo2 against log wet weight (A), CTMax (B), and Pcrit (C) values for each population of common killifish when acclimated to 20°C. All data are expressed as mean ± SEM, and significant differences between populations are indicated by different letters (p<0.001; F=16.223; df=2 for CTMax comparisons). Northern n= 19 (residual Mo2), 17 (CTMax), 15 (Pcrit); New Jersey n= 57 (residual Mo2), 54 (CTMax), 55 (Pcrit); Southern n= 16 (residual Mo2), 13 (CTMax), 14 (Pcrit).  36  Table 2. Mass specific metabolic rates of Northern (n= 19), New Jersey (n=57), and Southern (n=16) fish acclimated to 20°C Group: Mass specific metabolic rate (umol/g/hr) ± SEM Northern 9.631 ± 0.704 New Jersey 9.074 ± 0.426 Southern 7.888 ± 0.524  There was a significant relationship between the residuals of Mo2 and CTMax (p=0.00102, r= -0.354; Figure 3) when all fish were included in the analysis. There was also a significant correlation between residuals of Mo2 and Pcrit (p=2.22x10 -7 , r=0.535; Figure 4). In contrast, there was no significant relationship between the CTMax and Pcrit (p=0.666, r= -0.0517; Figure 5) 37   Figure 3. Inverse correlation between the residuals of metabolic rate, Mo2, and thermal tolerance, CTMax, in common killifish acclimated to 20°C. (Northern fish n=19; New Jersey n=63; Southern fish n=16; Pearson correlation: p=0.00102, r= -0.354) y = -0.4056x + 36.292 R 2 =0.125 p=0.00102 38        y = 44.379x + 37.546 R 2 =0.286 p<0.001 Figure 4. Significant positive correlation between residuals of metabolic rate, Mo2 and critical oxygen tension, Pcrit in common killifish acclimated to 20°C. (Northern fish n=14; New Jersey n=55; Southern fish n=14; Pearson correlation: p=2.2x10 -7 , r=0.535) 39        y = -0.0046x + 36.375 R 2 =2.67x10 -3  p=0.666 Figure 5. No significant correlation between critical oxygen tension, Pcrit, and thermal tolerance, CTMax, in common killifish acclimated to 20°C. (Northern fish n=13; New Jersey fish n=49; Southern fish n=10; Pearson correlation: p=0.666, r= -0.0517) 40  The effects of mitochondrial genotype and collection location on phenotype To examine the effects of mitotype and collection location on thermal tolerance and metabolic rate in fish from the contact zone, we divided fish from the contact zone (previously referred to as New Jersey fish) into three groups: fish from Red Bank with Northern mitotypes (Red Bank Northern), fish from Metedeconk with Northern mitotypes (Metedeconk Northern), and fish from Metedeconk with Southern mitotypes (Metedeconk Southern). None of the fish from Red Bank had Southern mitotypes. There was a significant difference (p=0.046; F=3.478; df=2) between the mean wet weights of contact zone groups. The mean wet weight of Red Bank Northern fish was 5.92 ± 0.416g, while Metedeconk Northern fish were 4.09 ± 0.364g and Metedeconk Southern fish were 5.138 ± 0.477g (mean ± SEM). As can be seen in Figure 6A, there was no significant difference between the mean CTMax values of Metedeconk fish carrying the different mitotypes (p=0.445; t=0.778; df=21). However, there was a significant difference in CTMax (p=0.003; t= -3.179; df=40) between individuals with the northern mitotype from Red Bank and Metedeconk (Figure 6B). Figure 7 summarizes the effect of mitochondrial genotype and collection location on the residuals of Mo2. Metedeconk Northern fish had a mean residual Mo2 of -0.104 ± 0.0425 while Metedeconk Southern fish had a mean residual Mo2 of 6.23x10 -3  ± 0.0531. There was a trend for a significant effect (p=0.0595; U=35; U’=97; n=11 Northern, 12 Southern; Figure 7A) of mitochondrial genotype in fish from Metedeconk as assessed using a Mann-Whitney U test that was used due to the Metedeconk Southern group failing the parametric assumption of normality. When we removed one individual from the Metedeconk Southern group who was almost 3 standard deviations from the mean, we obtained a significant difference (p=0.0192; U=25; U’=96; n=11 Northerns, 11 Southerns) between mitotype groups according to a Mann-Whitney 41  U test, which was run due to failing the parametric assumption of equal variances between groups. Similarly, there was a significant difference, according to a t-test, between these groups in the logarithm of mass-specific metabolic rate with the same individual removed (p=0.014; t=2.694; df=20).  The distribution of residuals of Mo2 for Metedeconk Northern, Metedeconk Southern and Red Bank Northern fish can be found in the Appendix (Figure A3). To determine whether there was an effect of sampling location on Mo2 (independent of mitotype), we compared the mean residual of Mo2 of fish with a Northern mitochondrial genotype from Red Bank (0.032 ± 0.018) to the Mo2’s of fish with a Northern mitochondrial genotype from Metedeconk. The Mann-Whitney U test was used because the assumption of equal variances was not met in either population and the Red Bank Northern group failed normality. There was no significant difference in Mo2 between these groups (p=0.9909; U=219; U’=221; n=37 Red Bank, 11 Metedeconk; Figure 7B). Mass-specific metabolic rates of each mitotype group within the two marshes are given in Table 3 to facilitate comparison with other studies.   Group Mass-specific RMR (umol/g/hr) Red Bank Northern 9.46 ± 0.48 Metedeconk Northern 9.77 ± 1.21 Metedeconk Southern 7.25 ± 0.95  Table 3. Mass-specific routine metabolic rates (mean ± SEM) for Red Bank fish (n=37), Metedeconk fish with Northern mitotypes (n=11) and Metedeconk fish with Southern mitotypes (n=12) acclimated to 20°C. 42   A B a a a b Figure 6. Mean CTMax values for groups by mitotype (A) and sample location (B). All data is presented as means ± SEM. Significant differences are represented by different letters in part B (p=0.003; t= -3.179; df=40). Metedeconk Northern n=11, Metedeconk Southern n=12, Red Bank n= 31. 43   Figure 7. Mean residual of Mo2’s for groups by mitochondrial genotype (A) and sample location (B). All data is presented as means ± SEM. (Red Bank Northern, n=37; Metedeconk Northern, n=11; Metedeconk Southern, n=12). A B a a a a 44  The effect of nuclear genotype and mismatch with mitochondrial genotype To examine the effects of nuclear genotype on metabolic rate, thermal tolerance and Pcrit, we genotyped all individuals at 7 nuclear microsatellite loci. There was no evidence that these microsatellite loci deviated from Hardy-Weinberg proportions and there was no indication of the presence of null alleles. We used these genotypes to estimate a hybrid index for each individual where a hybrid index of 1 is equivalent to a fully Northern nuclear genotype and a hybrid index of 0 is equivalent to a fully Southern nuclear genotype. As can be seen in Figure 8, there was no significant correlation between hybrid index and the residuals of Mo2 (p=0.287; Figure 8A), CTMax (p=0.730; Figure 8B), or Pcrit (p=0.697; Figure 8C). In order to determine whether there was evidence for an effect of a mismatch between mitochondrial and nuclear genotype on phenotype, we examined individuals from the hybrid zone bearing a Northern mitotype. There were too few individuals in our hybrid zone sample that bore the Southern mitotype (n=12) for us to make the reciprocal comparison. Figure 9 shows the residual of Mo2 (Figure 9A), CTMax (Figure 9B), and Pcrit (Figure 9C) for individuals with a Northern mitochondrial genotype plotted against nuclear hybrid index. Nuclear-mitochondrial incompatibilities would be expected to result in phenotypic differences between individuals with a Northern mitotype in a Northern nuclear background (H.I.=1) compared to individuals with a Northern mitochondrial genotype in a Southern nuclear background (H.I.=0). We saw no evidence of such an effect. 45   Figure 8. The insignificant correlations between residuals of Mo2 (A), CTMax (B), and Pcrit (C) and hybrid index of common killifish at 20°C. Metedeconk Northern n=11 (Mo2 and CTMax) and 8 (Pcrit); Metedeconk Southern n= 12; Red Bank Northern n= 35 (Mo2) and 31 (CTMax and Pcrit). A B C y = -0.0032x + 0.093 R2=0.0144 p=0.373 y = -0.0533x + 36.041 R2=5.20x10-4 p=0.870 y = 0.0468x + 35.995 R2=0.0199 p=0.325 46                   Figure 9. The range of residual Mo2 (A), CTMax (B), and Pcrit (C) values of fish from Red Bank and Metedeconk with Northern mitotypes and varying nuclear genotype (hybrid index) values. Individuals to the left of the graph are more similar to Southern parents, whereas individuals on the right are more similar to Northern parents. A B C 47  Testing the correlation between CTMax and Mo2 in fish from the contact zone To determine if there is a correlation between thermal tolerance and metabolic rate we examined this correlation in fish of mixed genetic ancestry from the contact zone. There was a significant correlation between CTMax and the residuals of Mo2 (p=0.0027; Figure 10), but predictive power was weak. 48  Figure 10. The significant inverse correlation between residuals of Mo2 and CTMax in fish from New Jersey (p=0.0192) acclimated to 20°C. (Red Bank Northern, n=31; Metedeconk Northern, n=11; Metedeconk Southern, n=12) y = -2.1421x + 35.988 R 2 =0.16 p=0.0027 49  DISCUSSION- The effect of body size on physiological phenotype  Variation in body size explained nearly half the variation in metabolic rate between individuals (R 2 =0.475), but was not significantly correlated with thermal tolerance or Pcrit (Figure 1). The positive correlation between wet weight and metabolic rate is not surprising considering that allometric scaling effects have been identified in previous work. Allometric scaling is the disproportionate increase or decrease in certain properties or characteristics of an animal’s morphology or physiology with an increase in size. In 1889, Rubner first proposed that under standardized conditions, larger animals will tend to have larger total metabolic rates since metabolic heat production is proportional to the surface area of an animal suggesting that the metabolic rate of an animal should be proportional to 2/3 the power of its body mass. Subsequently, Kleiber (1932) observed that the metabolic rate of an animal was proportional to the 3/4 power of its mass. However, the value of the exponent, as well as whether or not it is universal across taxa, is still under debate (Bokma 2004). My data suggest a scaling exponent of 0.834 which is significantly different from both of these values. However, Bokma (2004) calculated a scaling exponent of 0.86 in Sea Trout (Salmo trutta trutta) ranging from 0.1 to 600 g, which suggests that a higher scaling exponent may be normal for teleost fish. Regardless, previous studies have identified allometric scaling effects on metabolic rate in teleost fish (Clarke and Johnston 1999; Killen et al. 2007). Therefore, when comparing metabolic rates between groups, or even individuals, it is important to account for these affects in order to determine whether intra or interspecific differences are a result of differences in physiology or simply body mass. It is possible that since heat transfer between an object and its surroundings is related to the surface area of the object, CTMax may be correlated with body size. No significant 50  correlations were found, however, indicating that perhaps the difference in size among individuals tested wasn’t large enough to affect the rate of heat transfer or that this parameter is unimportant for Fundulus. There was also no significant correlation between Pcrit and wet weight, which was surprising considering that gill surface area is known to scale with body size in fish (Oikawa and Itazawa 1985). Pcrit, by definition, is the point at which an organism can no longer extract sufficient levels of oxygen from its environment to meet the demand of its cellular processes (McNab 2002). Since animals with larger gill surface areas may be able to extract more oxygen from their environment, they may exhibit lower critical oxygen tensions. My data, however, suggest that there is no correlation between body size and Pcrit. Physiology of New Jersey fish  Since wet weight was found to be positively correlated with metabolic rate, it was necessary to account for the effects of body size in our analyses of metabolic rate differences between populations. To compensate for the effects of body size on Mo2, we calculated residuals of the relationship between the logarithm of Mo2 and the logarithm of wet weight so that the difference between the predicted Mo2, based on wet weight, and the experimentally measured Mo2 represented a deviation from expectations. By comparing the mean deviation from predicted values between groups, we were able to detect differences in Mo2 that were not caused by differences in body size.  Fish from New Jersey had an intermediate mean residual Mo2 value which was very close to zero. This may be due to the fact that 60 of the fish included in this analysis were from New Jersey, which was far greater than the 18 Northern fish and 16 Southern fish that were analyzed; therefore New Jersey fish had a greater influence on the regression. Northern fish had a positive mean residual value, meaning that their measured Mo2’s were on average higher than 51  would be predicted for their body size; Southern fish had a negative mean residual value. However, none of these differences were significant. Despite the fact that no significant differences were found, the trend seen in this study in metabolic rate agrees with the findings of Fangue et al. (2009a). She found that Northern fish had a higher mass-specific whole animal metabolic rate than Southern fish across three different acclimation temperatures. Table 2 shows that the mean mass-specific metabolic of Northern fish and the New Jersey fish do not overlap with Southern fish. Although these differences were insignificant, the groups vary in the same direction as was found by Fangue et al. (2009a), supporting Fangue’s conclusion that the cold-adapted Northern subspecies have a higher metabolic rate than the warm-adapted Southern subspecies.  Killifish are an extremely thermal tolerant species, as can be seen in Figure 2B. Although all fish were acclimated to 20°C, Northern fish and New Jersey fish, both groups having significantly different thermal tolerance limits than the Southern fish, were able to tolerate 36.2°C and 36.0°C, respectively, while Southern fish could tolerate 37.6°C. The 1.4°C difference between the Northern and Southern fish from this study closely matches the 1.5°C difference Fangue et al. (2006) found when comparing the thermal tolerance limits of the same subspecies. Although this intraspecific difference in upper thermal tolerance limits is relatively small compared to the wide range of temperatures killifish are capable of acclimating to (Fangue et al. 2006), the difference may be enough to allow the subspecies to inhabit different thermal microhabitats. For example, Southern fish may be able to live in shallower marshes, where temperatures are warmer, than can Northern fish and may be able to move more freely between marshes when the tide is low. 52   The consistent differences in thermal tolerance between the subspecies, in my data as well as Fangue et al. (2006), suggests that the critical thermal tolerance limits of F. heteroclitus may be an intrinsic characteristic of the subspecies. However, without studies on laboratory- reared fish, it is impossible to determine if this characteristic is set by environmental factors, such as the thermal history of each subspecies, or genetic factors, such as mitochondrial or nuclear genotype. The variation in killifish thermal tolerance limits, however, does follow the predicted pattern based on average habitat temperatures. The fact that Southern fish have a higher upper thermal tolerance limit than Northern fish suggests that these subspecies may have undergone local adaptation to temperature. These whole-organism differences in killifish thermal biology provide support for the idea proposed by Powers and colleagues that differences in physiological specializations and genetic variation between subspecies most likely reflect adaptive responses to temperature, or other factors that could be correlated with latitude (Powers and Schulte 1998; Schulte 2001).  If the observed thermal tolerance differences between killifish subspecies were in fact a result of adaptation to varying environmental temperatures with latitude, we would expect that fish from New Jersey, being intermediate in latitude, would have intermediate thermal tolerance values. However, New Jersey fish had a thermal tolerance limit that was not significantly different from that of Northern fish, but significantly lower than that of Southern fish. This may be due to the fact that fish from Red Bank, NJ had a CTMax of 35.6°C, which was significantly lower than fish from New Hampshire (36.2°C), North Carolina (37.5°C), or Georgia (37.8°C). Therefore, when grouping all the New Jersey fish together, the Red Bank fish lowered the mean CTMax due to their larger sample size. Fish from Metedeconk, NJ, on the other hand, did have an intermediate CTMax (36.6°C), although this value was only significantly different from the 53  Southern populations (North Carolina and Georgia). From this data, we can see that, with the exception of fish from Red Bank, CTMax values increased with decreasing latitude, therefore, this data supports the theory that adaptive responses to temperature have led to thermal tolerance differences between the subspecies.  While differences in thermal tolerance were seen between the sub-species, differences in hypoxia tolerance, as measured by Pcrit, were harder to detect. Although none of the differences in Pcrit between the groups were significant, Southern fish appear to have a higher Pcrit than both Northern fish and New Jersey fish, with the New Jersey fish being intermediate between the two subspecies. One of the reasons that differences in Pcrit may be difficult to detect is that Pcrit was extremely variable, even within a group.  Pcrit’s of Northern fish ranged from 21.2 - 47.7 torr, New Jersey fish ranged from 16.7 - 89.8 torr, and Southern fish ranged from 28.0 - 57.0 torr. While the largest variability was found in the New Jersey fish, all of the groups encompassed a wide range of critical oxygen tensions. This enormous variability within groups makes it inherently difficult to detect statistical differences between groups. In addition to large variability, some fish exhibited patterns of oxy-conformation. These fish had metabolic rates that were dependent on environmental oxygen concentration throughout their trial. While these fish were few in number (4 fish in total), it raises questions about the overall strategy killifish employ during hypoxia and whether or not killifish are even true oxy- regulators. Further study will be needed to try to determine a cause for the pattern of oxy- conformation seen in these rare individuals. Richards et al. (2008) measured a Pcrit of 63.9 ± 5 torr (n=6) for Northern killifish acclimated to 20°C. This is drastically different from the Pcrit value of 34.487 ± 1.987 torr (n=15) 54  we obtained for Northern fish. One explanation could be the fact that this experiment used more fish than previous work. As discussed above, there are some anomalies in the metabolic strategies of killifish exposed to hypoxia; therefore, it is possible that measuring a smaller number of fish could lead to increased variability and, perhaps, decreased accuracy in Pcrit measurements. In addition, we calculated our Pcrit values using the measured routine metabolic rate for each fish and then regressing through the data points that were more than 12% below the routine metabolic rate, whereas Richards et al. used a breakpoint regression program (Yeager and Ultsch 1989). This program determines the intersection point between two segmented regression lines to determine Pcrit. It is possible that because our method only regressed through points that were below the RMR that we may have underestimated the Pcrit slightly. However, this method, in our opinion, is more accurate as it accounts for the variability that is inherent with measuring metabolic rates in this species. The observation that Southern fish tended (nonsignificantly) to have the highest Pcrit’s was opposite of our prediction that Southern fish should have a lower Pcrit than Northerns based on their higher Hb-oxygen affinity. Previous work has shown that fish possessing the LDH-B SS  genotype, which is fixed in Southern populations, have a lower ATP/Hb ratio and therefore a higher Hb-oxygen affinity (Powers et al. 1979). This led Powers and colleagues (1979) to predict that Southern fish should be more competitive in hypoxic environments, and thus, our prediction that Southern fish should have a lower critical oxygen tension. However, it is possible that having a higher Hb-O2 affinity means that oxygen is less likely to be released at the tissues. If metabolizing tissues fail to receive sufficient oxygen, this will increase the likelihood that the animal will decrease its metabolic rate in order to bring oxygen demand to a level that can be matched by the supply available. Thus, Southern fish, because of their increased affinity of 55  haemoglobin for oxygen, may actually be less efficient at delivering oxygen to tissue and, thus, may be less hypoxia tolerant than the Northern subspecies. Oxygen-limited thermal tolerance across the species range  The Oxygen-Limited Thermal Tolerance (OLTT) hypothesis states that an organism’s thermal tolerance is linked to the point at which it is unable to match oxygen supply with oxygen demand (Pörtner 2002). This is seen through a decrease in aerobic scope, which is the organism’s maximal metabolic rate minus its routine metabolic rate. As temperatures increase past the optimal temperature, routine metabolic rate increases faster than maximal metabolic rate, thus a reduction in aerobic scope occurs. Therefore, an organism’s routine metabolic rate could be an important factor in setting upper thermal tolerance limits (CTMax) due to its effect on aerobic scope. The higher an organism’s routine metabolic rate, the lower its aerobic scope and, thus, the lower its thermal tolerance will be (Pörtner 2001).  In order to test the validity of the OLTT hypothesis, we examined the relationship between CTMax and metabolic rate in the common killifish. As shown in Figure 3, the residuals of Mo2 and CTMax were inversely correlated (p<0.001) when a single linear regression was run through all individuals that were measured. This inverse correlation supports Pörtner’s idea that oxygen limitations may actually be the driving force setting thermal limits of organisms. Because oxygen solubility decreases as water temperature increases, oxygen supply decreases as temperature increases, which would place strain on the cardiovascular and respiratory systems of the individual, even if metabolic rate did not increase with temperature. The fact that organisms with higher metabolic rates feel the effects of heat stress sooner than organisms with lower metabolic rates suggests that maintaining a higher metabolic rate in rising temperatures is more stressful, perhaps due to the decreased availability of oxygen in the environment. 56   This inverse correlation, although not directly tested, was observed by Fangue et al. (2009a).  She found that Northern fish had significantly lower upper thermal tolerance limits, by approximately 1.5°C, than Southern fish across a wide range of acclimation temperatures (Fangue et al. 2006). In previous work, she found that the Northern fish had a significantly higher routine metabolic rate than Southern fish across three different acclimation temperatures. Therefore the fish with the higher metabolic rate (Northerns) also had the lowest thermal tolerance. However, Fangue et al. (2009a) could not attribute these whole-animal differences to adjustments in mitochondrial amount of function in warm acclimation temperatures. Therefore, it appears that the mechanisms through which organisms deal with oxygen limitations with increasing temperatures are complex. However, whole-animal metabolism may be the most accurate way to determine when an organism will become oxygen limited as it represents the total demand of the organism.  Pörtner’s theory of oxygen limited thermal tolerance suggests that fish with a higher metabolic rate will become oxygen-limited more quickly (Pörtner 2001). Therefore, fish with a higher metabolic rate might also show that they are less tolerant to decreases in environmental oxygen tensions and, as a result, will have a higher critical oxygen tension. Figure 4 shows that there is in fact a significant (p<0.001) positive correlation between the residuals of routine metabolic rate and Pcrit. From this correlation, it can be suggested that fish with a higher metabolic rate experienced a mismatch between oxygen supply and demand more quickly than fish with lower metabolic rates, and therefore reduced their routine oxygen demand at higher oxygen tensions. While Pcrit is a proxy for hypoxia tolerance and not thermal tolerance, this finding does lend support to Pörtner’s theory that having a higher metabolic rate makes a fish more susceptible to oxygen limitation. 57   Since oxygen limitation is the proposed mechanism for thermal limits, a more hypoxia tolerant fish may exhibit greater thermal tolerance as well. Figure 5 shows that there was no significant correlation (p=0.666) between CTMax and Pcrit, however. One explanation for this may be that thermal tolerance is more dependent upon blood oxygen levels (Pörtner 2002), whereas critical oxygen tension measures the environmental oxygen level at which the organism becomes limited. Therefore an organism may have adapted a larger gill surface area, for example, to extract more oxygen from the environment, but, when temperatures increase metabolic rate may increase rapidly, resulting in an overall decrease in hypoxia of bodily fluids, thereby decreasing oxygen supply to the tissues. In our determination of critical oxygen tension, fish were held at a constant temperature; therefore no increase in metabolic rate occurred. This means that any adaptations the subspecies possess for surviving environmental hypoxia would increase the supply to tissues. During thermal stress, however, the demand for oxygen increases to a point where the circulatory system cannot supply oxygen fast enough (Pörtner 2002). For this reason, physiological adaptations that result in increased hypoxia tolerance may not serve to increase thermal tolerance. New Jersey fish  Because the observed correlations could be driven by the fish from the two extreme populations, we also examined these relationships in fish from New Jersey alone. We reasoned that if the correlation between CTMax and Mo2 was a functional constraint, then it should also be present in New Jersey fish alone. We also examined the effects of mitochondrial genotype, collection location and nuclear genotype on the thermal tolerance and metabolic rate of all individuals from New Jersey and discuss the validity of the OLTT hypothesis with respect to hybrid individuals. 58  Effect of collection location and mitochondrial genome on physiology  In order to determine the effects of mitochondrial genotype on thermal tolerance, we compared fish from the same sample location, but with different mitochondrial haplotypes. We found no significant difference in CTMax between fish from the Metedeconk marsh with a Northern mitotype and fish from the same marsh with a Southern mitotype (p=0.445; t=0.778; df=21; Figure 6A). We were unable to make the same comparison for the Red Bank marsh since all the fish that were examined were found to have the Northern mitotype. These results suggest that the mitochondrial genome itself likely does not determine the thermal tolerance of an organism. What is more likely, however, is that mitochondrial properties controlled by the nuclear genome dictate the upper thermal limits of an organism. For example, Fangue et al. (2009a) found that cold-acclimated Northern fish had higher mitochondrial enzyme activities and mRNA levels, which correlated with higher mass-specific metabolic rates, than Southern fish acclimated to the same temperature. Her results suggested that the putatively cold-adapted Northern subspecies may even have a greater ability to increase mitochondrial numbers in response to cold-acclimation (Fangue et al. 2009a), which would serve to increase the whole- animal metabolic rate of the subspecies. Therefore it seems that the function and the amount of mitochondria, not the genes themselves, are important for setting differences in metabolic rate. While we did not see a correlation between mitochondrial genotype and CTMax, we did, however, see a pattern in thermal tolerance values when we examined fish from different sample locations having the same mitochondrial genotype. When comparing fish with a Northern mitochondrial genotype from Red Bank and Metedeconk, we saw a significant difference (p=0.003; t=-3.179; df=40) in mean CTMax values between locations with fish from Metedeconk, the more Southerly location, having the highest thermal tolerance (Figure 6B). Red 59  Bank fish had a CTMax of 35.6°C, whereas fish from Metedeconk had a CTMax of 36.4°C. The same comparison could not be made for fish with Southern mitotypes, since none of the fish from the Red Bank location had a Southern mitochondrial genome. However, it is hard to imagine that a significant effect of collection location would apply only to fish of one mitochondrial genotype. Presumably, location effects would affect fish with both mitochondrial genomes equally. The fact that location seems to be a better predictor of thermal tolerance than mitochondrial genotype suggests that an organism’s thermal history could be important in setting CTMax values. An organism which has been exposed to higher temperatures at some point in its life history may be better prepared to tolerate high temperatures at a later stage (Somero 2002). In other words, occasional exposure to high, but non-lethal temperatures may help increase an organism’s thermal tolerance. This exposure-induced tolerance may be possible through the induction of heat shock proteins (hsps). Hsps are chaperone proteins that help re-fold proteins which have been denatured through various stressors, such as hypoxia, temperature and salinity (Feder and Hofmann 1999). Evidence for this is seen in fruit flies which show an increased frequency of survival to heat stress after heat hardening at a non-lethal temperature has occurred (Sejerkilde et al. 2003). Drosophila  that were exposed to thermal stress also showed increased amounts of Hsp70, but the response was delayed until after the organism was returned to a permissive temperature (Sejerkilde et al. 2003). The delayed elevation of hsp expression in response to thermal stress may be what allows organisms to tolerate thermal stress experienced later in life. Sejerkilde et al.’s results suggest that organisms in a hypervariable environment, such as animals living in an intertidal zone, may be more tolerant to various stressors, such as heat and oxygen, as a result of elevated hsp levels. 60  Although thermal tolerance was found to be correlated with collection location, metabolic rate showed no such correlation. When fish with a Northern mitotype from either Red Bank or Metedeconk were compared, they had metabolic rates that were not significantly different (p=0.9909; U=219; U’=221; n= 37 Red Bank, 11 Metedeconk; Figure 7B). However, when fish from Metedeconk with different mitochondrial genomes were compared, we saw a trend towards a significant difference in the residuals of Mo2 between fish with a Northern mitotype and fish with a Southern mitotype (p=0.0595; U=35; U’=97; n=11 Northern, 12 Southern; Figure 7A). However, one individual in the Metedeconk Southern group was nearly three standard deviations from the mean of the group. When this individual’s residual Mo2 was removed from analysis, we saw a significant difference (p=0.0192; U=25; U’=96; n=11 Northern, 11 Southern) between the Northern and Southern mitotypes in the Metedeconk marsh. Taken together, these two analyses strongly suggest that there is a significant difference in Mo2 which is correlated with mitochondrial genotype. Since Northern fish have higher whole-animal routine metabolic rates than Southern fish across a variety of temperatures (Fangue et al. 2009a), one could predict that fish with a Northern mitochondrial genotype would have slightly higher routine Mo2’s. The fact that we do see a correlation between mitochondrial genotype and routine metabolic rate is not surprising given the genes that are encoded by the mitochondrial genome. Perhaps the most important of these genes are the ones that code for proteins in the electron transport chain (ETC). Thirteen of the proteins encoded by the mitochondrial genome are components of the ETC, and, therefore, directly affect the oxygen consumption of a cell (Blier et al. 2001; Burton et al. 2006). Depending on the mitochondrial genotype of a fish, different isoforms of various ETC enzymes could be expressed which could result in different activity levels between mitotypes. Evidence 61  for differing activity levels of ETC enzymes between the subspecies is mixed, however.  Fangue et al. (2009a) found when acclimated to 5°C, Northern fish had significantly higher citrate synthase and cytochrome c oxidase activity levels in some tissues. However, this difference became insignificant at higher acclimation temperatures. Nevertheless, these results, combined with our findings that metabolic rate was significantly different in fish with different mitotypes, suggest that differences at the mitochondrial genome level may be responsible for differences in physiology between fishes. Despite the fact that thermal tolerance seems to be correlated with collection location and not mitochondrial genotype, it is difficult to eliminate genotype as a determining factor until lab- reared fish can be analyzed. Likewise, it is difficult to say that collection location is a poor predictor of metabolic rate unless the environment of an organism is known throughout its life. While all fish used in these experiments were caught at the same time of year and were all from marshes with similar dissolved oxygen and salinity the temperature of the Metedeconk marsh during deployment of the trap was 20.4°C, while temperature at collection was 27.4°C. Although both Northern and Southern mitotype fish were caught at the Metedeconk location, one mitotype of fish may have been trapped earlier in the day when environmental conditions were slightly different. It is, therefore, impossible to tell whether or not there was a difference in temperature when fish of each mitochondrial genome were caught as well as whether or not the conditions in which the fish were caught are representative of their typical habitat. In order to verify the results seen in wild-caught hybrid fish, common-garden experiments could be run on lab-reared hybrids to control for all variables other than genotype. By controlling the thermal history, dissolved oxygen and salinity (among other variables) each fish is exposed to, one could more confidently 62  attribute differences in thermal tolerance and metabolic rate to genotype rather than life history traits. Effect of nuclear genotype on phenotype  One of the goals of this thesis was to determine whether or not nuclear genotype was a good predictor of physiological phenotype. We found, however, that nuclear genotype, as assessed using microsatellites, does not appear to be a good predictor of Mo2, CTMax, or Pcrit. When all the New Jersey fish were included in a linear regression analysis, we obtained nearly horizontal trend lines, all of which were insignificant (Figure 8). Therefore, there does not appear to be any correlation between nuclear genotype, as estimated by microsatellite markers, and physiological phenotype. In generating our hybrid index, we chose to use putatively neutral microsatellite markers to generate a hybrid index as an estimation of nuclear genotype. The benefit of using neutral markers lies in the ability to detect any effects of genetic variation at lesser-known loci (Dalziel et al. 2009). Because microsatellites are neutral, unless they are linked to other loci which are selected for, they are able to pass freely between subspecies. As a result, we did not see any fixed differences between subspecies at any of the microsatellites used in this analysis (J. McKenzie, unpublished data). However, the hybrid index values did follow a clinal pattern when fish from multiple sites throughout the contact zone were analyzed, with individuals with Southern nuclear genotypes occurring with greater frequency in more Southern locations (J. McKenzie, unpublished data). In the future, it would be interesting to use additional microsatellite markers to get an even more precise calculation of hybrid index. Analysis of additional microsatellites might 63  reveal some neutral markers with fixed differences between the subspecies, increasing the accuracy of a hybrid index calculation. In addition, it would be interesting to compare our results to a study that calculates hybrid index based on markers which are under selection, particularly if the selected markers were involved in processes that could affect the physiology of the fish. For example, using the genotype of LDH-B, a glycolytic enzyme, to generate a hybrid index would give a good indication of how the genotype of an individual for a gene directly involved in physiological processes affects the physiology of an individual. Cytonuclear incompatibility  In order to determine if there were any effects of mismatch between the mitochondrial and nuclear genome on the physiology of hybrid fish, we analyzed hybrid individuals, from either Red Bank or Metedeconk, with a Northern mitotype. If there was an effect of mismatch between the genomes on these hybrid fish, we would expect that Northern mitotype fish with a Southern nuclear genotype, indicated by hybrid index values close to zero, would have physiological phenotypes that were significantly different from Northern mitotype fish with a Northern nuclear genotype. However, we did not see any such trend in our data (Figure 9). Northern mitotype fish with a hybrid index of 1 (fully Northern nuclear genotype), for example, had mean residuals of Mo2, CTMax, and Pcrit that were similar to Northern mitotype fish with a hybrid index of zero (fully Southern nuclear genotype). The effects of mismatch between the genomes in fish with Southern mitotypes couldn’t be analyzed since none of the fish from Red Bank were found to have a Southern mitotype and only 12 fish from Metedeconk possessed a Southern mitochondrial genome.   The fact that we did not see a physiological effect of mismatch between the nuclear and mitochondrial genotypes suggests that two possibilities: either there is no effect of cytonuclear 64  incompatibilities on physiology, or the nuclear markers that were used were not able to detect an effect of mismatch since they do not appear to be linked to any genes of physiological importance.  In order to distinguish between these two possibilities, it would be interesting to genotype the hybrid fish for proteins involved in the ETC and compare the physiological phenotypes of fish with mismatches between nuclear-encoded components of the ETC and mitochondrial genotype. For example, J. McKenzie (unpublished data) has found a fixed difference between the subspecies in one of the nuclear-encoded subunits of the NADH dehydrogenase complex, which is an enzyme in the inner mitochondrial membrane that is responsible for passing electrons through the ETC and, ultimately, leading to ATP production. Examining the physiological response of mismatch between mitochondrial genotype and genotype at this nuclear-encoded ETC enzyme may give a better indication of the effects of cytonuclear incompatibilities. Oxygen-limited thermal tolerance in hybrids?  As we found in our analysis of all fish, including those from the extremes of the species range, we saw a significant inverse correlation (p=0.0027; Figure 10) between the residuals of Mo2 and CTMax when analyzing fish from New Jersey alone. The R 2  value of the correlation suggests that approximately 16% of the individual variation in CTMax could be explained by variation in the individual’s routine metabolic rate. Despite having intermediate CTMax and residual Mo2 values compared to fish from the latitudinal extremes, the Mo2 of fish from the contact zone still proves to be a good predictor of CTMax at an individual level.  This finding further supports Pörtner’s hypothesis that oxygen limitation is the driving force behind thermal limits (Pörtner 2001). Hybrid individuals with higher Mo2’s were less capable of tolerating increasing temperatures than individuals with lower metabolic rates. In 65  order to provide more support for the idea that it is actually oxygen limitation that limits the thermal tolerance of an organism, future work should focus on assessing the routine and maximal metabolic rates of animals as temperature increases towards their CTMax. This will give a better indication of whether or not decreases in aerobic scope are causally related to decreased thermal tolerance. 66  CONCLUSIONS  Overall, my research shows that there does appear to be an inverse correlation between routine metabolic rate and thermal tolerance, as was first suggested by the oxygen-limited thermal tolerance hypothesis. From data presented in this thesis, it appears that increasing temperature may create a mismatch between oxygen supply and demand and that organisms with a higher baseline demand for oxygen are less tolerant of increasing temperatures. Therefore, organisms that are able to extract and utilize more oxygen from the environment should be more tolerant of high temperatures.  In addition, we found that collection location was strongly correlated with thermal tolerance, suggesting that the thermal history of an organism determines the temperatures it can tolerate. This is likely due to phenotypic plasticity where prior environmental stressors enhance the stress response of the organism when exposed to the same stress or a different stress later in its life. There was no effect of collection location on routine metabolic rate.  Despite the lack of correlation between nuclear genotype and any of the three physiological variables measured, we did see a correlation between mitochondrial genotype and routine metabolic rate. There was a trend towards significance for fish with Northern mitochondrial genotypes to have a higher routine metabolic rate than fish with Southern mitochondrial genotypes. We did not see any effect of mismatch between the nuclear and mitochondrial genomes on physiological properties. 67  BIBLIOGRAPHY Able, KW. and J.D. Felley. 1986. Geographical Variation in Fundulus heteroclitus: Tests for Concordance between Egg and Adult Morphologies. American Zoologist 26(1):145-157.  Abraham, B.J. 1985. Species Profiles: Life Histories and Environmental Requirements of Coastal Fishes and Invertebrates (Mid-Atlantic): Mummichog and Striped Killifish. U.S. Fish and Wildlife Services Biological Report 82(11.40). U.S. Army Corps of Engineers, TR EL-82-4. 23 pp.  Adams, S.M., M.F. Oleksiak, and D.D. Duvernall. 2005. Microsatellite primers for the Atlantic coastal killifish, Fundulus heteroclitus, with applicability to related Fundulus species. Molecular Ecology 5:275-277.  Adams, S.M., J.B. Lindmeier, and D.D. Duvernall. 2006. Microsatellite analysis of the phylogeography, Pleistocene history and secondary contact hypotheses for the killifish, Fundulus heteroclitus. Molecular Ecology 15:1109-1123.  Alabaster, J.S., and R.L. Welcomme. 1962. Effect of Concentration of Dissolved Oxygen on Survival of Trout and Roach in Lethal Temperatures. Nature 4823:107.  Atz, J.W. 1986. Fundulus heteroclitus in the laboratory: A History. American Zoologist 26:111- 120.  Becker, C.D., and R.G. Genoway. 1979. Evaluation of the critical thermal maximum for determining thermal tolerance of freshwater fish. Environmental Biology of Fishes 4(3):245-256.  Beitinger, T.L., W.A. Bennett, and R.W. McCauley. 2000. Temperature tolerances of North American freshwater fishes exposed to dynamic changes in temperature. Environmental Biology of Fishes 58:237-275.  Bernardi, G., P. Sordino, and D.A. Powers. 1993. Concordant mitochondrial and nuclear DNA phylogenies for populations of the teleost fish Fundulus heteroclitus. Proceedings of the National Academy of Science 90:9271-9274.  Blier, P.U., F. Dufresne, and R.S. Burton. 2001. Natural selection and the evolution of mtDNA- encoded peptides: evidence for intergenomic co-adaptation. TRENDS in Genetics 17(7):400- 406.  Bokma, F. 2004. Evidence against universal metabolic allometry. Functional Ecology 18:184- 187.  Boutilier, R., T.A. Heming, and G.K. Iwama. Appendix: physicochemical parameters for use in fish respiratory physiology. In: Fish Physiology, edited by W.S. Hoar and D.J. Randall. New York: Academic, 1984, p. 403-430.  68  Brown, B.L. and R.W. Chapman. 1991. Gene Flow and Mitochondrial DNA Variation in the Killifish, Fundulus heteroclitus. Evolution 45(5):1147-1161.  Brown, W.M., M. George, and A.C. Wilson. 1979. Rapid Evolution of Animal Mitochondrial DNA. Proceedings for the National Academy of Science of the United States of America 76(4):1967-1971.  Burton, R.S. 1990. Hybrid Breakdown in Developmental Time in the Copepod Tigriopus californicus. Evolution 44(7):1814-1822.  Burton, R.S., C.K. Ellison, and J.S. Harrison. 2006. The Sorry State of F2 Hybrids: Consequences of Rapid Mitochondrial DNA Evolution in Allopatric Populations. The American Naturalist 168:S14-S24.  Chapman, L.J., C.A. Chapman, F.G. Nordlie, and A.E. Rosenberger. 2002. Physiological refugia: swamps, hypoxia tolerance and maintenance of fish diversity in the Lake Victoria region. Comparative Biochemistry and Physiology Part A 133:421-437.  Clarke, A. and N.M. Johnston. 1999. Scaling of metabolic rate with body mass and temperature in teleost fish. Journal of Animal Ecology 68:893-905.  Cox, D.K. 1974. Effects of three heating rates on the critical thermal maximum of Bluegill. In: Thermal Ecology (ed. W. Gibbons and R.R. Sharitz), pp.158-163. Springfield, IL: National Technical Information Service.  Dalziel, A.C., S.M. Rogers, and P.M. Schulte. 2009. Linking genotypes to phenotypes and fitness: how mechanistic biology can inform molecular ecology. Molecular Ecology 18: 4997- 5017.  DiMichele, L. and D.A. Powers. 1982a. LDH-B genotype-specific hatching times of Fundulus heteroclitus embryos. Nature 296:563-564.  DiMichele, L. and D.A. Powers. 1982b. Physiological Basis for Swimming Endurance Differences between LDH-B Genotypes of Fundulus heteroclitus. Science 216(4519):1014- 1016.  DiMichele, L. and D.A. Powers. 1984. Developmental and Oxygen Consumption Rate Differences between Lactate Dehydrogenase-B Genotypes of Fundulus heteroclitus and Their Effect on Hatching Time. Physiological Zoology 57(1):52-56.  DiMichele, L., D.A. Powers, and J.A. DiMichele. 1986. Developmental and Physiological Consequences of Genetic Variation at Enzyme Synthesizing Loci in Fundulus heteroclitus. American Zoologist 26:201-208.  DiMichele, L., K.T. Paynter, and D.A. Powers. 1991. Evidence of Lactate Dehydrogenase-B Allozyme Effects in the Teleost, Fundulus heteroclitus. Science 253(5022):898-900. 69   DiMichele, L. and M.E. Westermann. 1997. Geographic variation in development rate between populations of the teleost Fundulus heteroclitus. Marine Biology 128:1-7.  Duvernell, D.D., J.B. Lindmeier, K.E. Faust, and A. Whitehead. 2008. Relative influences of historical and contemporary forces shaping the distribution of genetic variation in the Atlantic killifish, Fundulus heteroclitus. Molecular Ecology 17:1344-1360.  Edmands, S. 1999. Heterosis and Outbreeding Depression in Interpopulation Crosses Spanning a Wide Range of Divergence. Evolution 53(6):1757-1768.  Edmands, S. and R.S. Burton. 1999. Cytochrome C Oxidase Activity in Interpopulation Hybrids of a Marine Copepod: A Test for Nuclear-Nuclear or Nuclear-Cytoplasmic Coadaptation. Evolution 53(6):1972-1978.  Ellison, C.K., and R.S. Burton. 2006. Disruption of mitochondrial function in intertidal hybrids of Tigriopus californicus. Evolution 60(7):1382-1391.  Ellison, C.K., and R.S. Burton. 2008. Interpopulation hybrid breakdown maps to the mitochondrial genome. Evolution 62(3):631-638.  Fangue, N.A., M. Hofmeister, and P.M. Schulte. 2006. Intraspecific variation in thermal tolerance and heat shock protein gene expression in common killifish, Fundulus heteroclitus. The Journal of Experimental Biology 209:2859-2872.  Fangue, N.A., J.G. Richards, and P.M. Schulte. 2009a. Do mitochondrial properties explain intraspecific variation in thermal tolerance? The Journal of Experimental Biology 212:514-522. Fangue, N.A., J.E. Podrabsky, L.I. Crawshaw, and P.M. Schulte. 2009b. Countergradient Variation in Temperature Preference in Populations of Killifish Fundulus heteroclitus. Physiological and Biochemical Zoology 82(6):776-786.  Feder, M.E. and G.E. Hoffman. Heat-shock proteins, molecular chaperones and the stress response: Evolutionary and Ecological physiology. Annual Review of Physiology 61:245-282.  Fritz, E.S., W.H. Meredith, and V.A. Lotrich. 1975. Fall and Winter Movements and Activity Levels of the Mummichog, Fundulus heteroclitus, in a Tidal Creek. Chesapeake Science 16(3):211-215.  Fry, F.E.J. 1947. Effects of the environment on animal activity. University of Toronto Studies Biological Series No. 55, Toronto: Publications of the Ontario Fisheries Research Laboratory 68: 1-62  Gompert, Z. and C.A. Buerkle. 2009. A powerful regression-based method for admixture mapping of isolation across the genome of hybrids. Molecular Ecology 18:1207-1224.  70  Gompert, Z. and C.A. Buerkle. 2010. INTROGRESS: a software package for mapping components of isolation in hybrids. Molecular Ecology Resources 10:378-384.  Gonzalez-Villasenor, L.I. and D.A. Powers. 1990. Mitochondrial-DNA Restriction-Site Polymorphisms in the Teleost Fundulus heteroclitus Support Secondary Intergradation. Evolution 44(1):27-37.  Grebmeier, J.M., J.E. Overland, S.E. Moore, E.V. Farley, E.C. Carmack, L.W. Cooper, K.E. Frey, J.H. Helle, F.A. McLaughlin, S.L. McNutt. 2006. A Major Ecosystem Shift in the Northern Bering Sea. Science 311:1461-1464. Hardy, J. D. Jr.  1978. Development of fishes of the mid-Atlantic bight: An atlas of egg, larval and juvenile stages. In Volume II Anguillidae through Syngnathidae, U.S. Department of the Interior, pp 141-200.  Henriksson, P., M. Mandic, and J.G. Richards. 2008. The Osmorespiratory Compromise in Sculpins: Impaired Gas Exchange Is Associated with Freshwater Tolerance. Physiological and Biochemical Zoology 81(3):310-319.  Hochachka, P.W. and G.N. Somero. 2002. Biochemical adaptation: mechanism and process in physiological evolution. Oxford University Press, New York.  Hughes, G.M. 1973. Respiratory Responses to Hypoxia in Fish. American Zoologist 13(2):475- 489.  Killen, S.S., I. Costa, J.A. Brown, and A.K. Gamperl. 2007. Little left in the tank: metabolic scaling in marine teleosts and its implications for aerobic scope. Proceedings of the Royal Society B 274:431-438.  King, M.P. and G, Attardi. 1989. Human Cells Lacking mtDNA: Repopulation with Exogenous Mitochondria by Complementation. Science 246(4929):500-503.  Kleiber, M. 1932. Body size and metabolism. Hilgardia 6:315-332.  Lane, N. 2009. On the Origin of Bar Codes. Nature 462:272-274.  Lee, W., J. Conroy, W. Hunting Howell, T.D. Kocher. 1995. Structure and Evolution of Teleost Mitochondrial Control Regions. Journal of Molecular Evolutions 41:54-66.  Liepins, A. and S. Hennen. 1977. Cytochrome Oxidase Deficiency during Development of Amphibian Nucleocytoplasmic Hybrids. Developmental Biology 57:284-292.  Lotrich, V.A. 1975. Summer Home Range and Movements of Fundulus Heteroclitus (Pisces: Cyprinodontidae) in a Tidal Creek. Ecology 56(1):191-198.  71  Mandic, M., A.E. Todgham, and J.G. Richards. 2009. Mechanisms and evolution of hypoxia tolerance in fish. Proceedings of the Royal Society B 276:735-744. McNab, B.K. 2002. The Physiological Ecology of Vertebrates: a View from Energetics. Cornell University Press, Ithaca. Mied, P.A., and D.A. Powers. 1978. Hemoglobins of the Killifish Fundulus heteroclitus. The Journal of Biological Chemistry 253(10): 3521-3528.  Mitton, J.B., and R.K. Koehn. 1975. Genetic organization and adaptive response of allozymes to ecological variables in Fundulus heteroclitus. Genetics 79:99-111.  Mitton, J.B., and R.K. Koehn. 1976. Morphological Adaptation to Thermal Stress in a Marine Fish, Fundulus heteroclitus. Biological Bulletin 151(3): 548-559.  Morin, R.P., and K.W. Able. 1983. Patterns of Geographic Variation in the Egg Morphology of the Fundulid Fish, Fundulus heteroclitus. Copeia 1983(3):726-740.  Moritz, C., T.E. Dowling, and W.M. Brown. 1987. Evolution of Animal Mitochondrial DNA: Relevance for Population Biology and Systematics. Annual Review of Ecology and Systematics 18:269-292.  Mousseau, T.A., and D.J. Howard. 1998. Genetic Variation in Cricket Calling Song Across a Hybrid Zone Between Two Sibling Species. Evolution 52(4):1104-1110.  Niehuis, O., A.K. Judson, and J. Gadau. 2008. Cytonuclear Genetic Incompatibilities Cause Increased Mortality in Male F2 Hybrids of Nasonia giraulti and N. vitripennis. Genetics 178:413- 426.  Oikawa, S. and Y. Itazawa. 1985. Gill and body surface areas of the carp in relation to body mass, with special reference to the metabolism-size relationship. Journal of Experimental Biology 117:1-14.  Paynter, K.T., L. DiMichele, S.C. Hand, and D.A. Powers. 1998. Metabolic Implications of Ldh- B genotype during early development in Fundulus heteroclitus. Comparative Physiology and Biochemistry 257(1):24-33.  Perry, A.L., P.J. Low, J.R. Ellis, and J.D. Reynolds. 2005. Climate Change and Distribution Shifts in Marine Fishes. Science 308:1912-1915. Place, A.R., and D.A. Powers. 1979. Genetic Variation and Relative Catalytic Efficiencies: Lactate Dehydrogenase B Allozymes of Fundulus heteroclitus. Proceedings of the National Academy of Science 76(5):2354-2358.  Place, A.R., and D.A. Powers. 1984. Kinetic Characterization of the Lactate Dehydrogenase (LDH-B4) Allozymes of Fundulus heteroclitus. The Journal of Biological Chemistry 259(2):1309-1318. 72   Pörtner H.O. 2001. Climate change and temperature-dependent biogeography: oxygen limitation of thermal tolerance in animals. Naturwissenschaften 88:137-146.  Pörtner H.O. 2002. Climate variations and the physiological basis of temperature dependent biogeography: systemic to molecular hierarchy of thermal tolerance in animals. Comparative Biochemistry and Physiology Part A 132:739-761. Pörtner H.O., and M.K. Grieshaber. Critical Po2(s) in oxyconforming and oxyregulating animals: gas exchange, metabolic rate and the mode of energy production. In: The Vertebrate Gas Transport Cascade: Adaptations to Environment and Mode of Life, edited by J. Eduardo and P.W. Bicudo. Boca Raton: CRC, 1993. Pörtner H.O., I. Hardewig, F.J. Sartoris, and P.L.M. van Dijk. 1998. Energetic aspects of cold adaptation: critical temperatures in metabolic, ionic and acid-base regulation? In: Cold Ocean Physiology. (ed. H.O. Pörtner and R. Playle). Pp.88-120. Cambridge University Press, Cambridge. Pörtner, H.O. and R. Knust. 2007. Climate change affects marine fishes through the oxygen limitation of thermal tolerance. Science 315:95-97.  Powers, D.A., and A.R. Place. 1978. Biochemical Genetics of Fundulus heteroclitus (L.). I. Temporal and Spatial Variation in Gene Frequencies of Ldh-B, Mdh-A, and Pgm-A 1 . Biochemical Genetics 16(5/6):593-607.  Powers, D.A., G.S. Greaney, and A.R. Place. 1979. Physiological correlation between lactate dehydrogenase genotype and haemoglobin function in killifish. Nature 277:240-241.  Powers, D.A., I. Ropson, D.C. Brown, R. Van Beneden, R. Cashon, L.I. Gonzalez-Villasenor, and J.A. DiMichele. 1986. Genetic Variation in Fundulus heteroclitus: Geographic Distribution. American Zoology 26:131-144.  Powers, D.A., T. Lauerman, D. Crawford, L. DiMichele. 1991. Genetic Mechanisms for Adapting to a Changing Environment. Annual Review of Genetics 25:629-659.  Powers, D.A., M. Smith, I. Gonzalez-Villasenor, L. DiMichele, D. Crawford, G. Bernardi, T. Lauerman. 1993. A multidisciplinary approach to the selectionist/neutralist controversy using the model teleost, Fundulus heteroclitus. Oxford Surveys in Evolutionary Biology 9:43-107.  Powers, D.A., and P.M. Schulte. 1998. Evolutionary Adaptations of Gene Structure and Expression in Natural Populations in Relation to a Changing Environment: A Multidisciplinary Approach to Address the Million-Year Saga of a Small Fish. The Journal of Experimental Zoology 282:71-94.  Pritchard, J.K., M. Stephens, and P. Donnelly. 2000. Inference of Population Structure Using Multilocus Genotype Data. Genetics 155:945-959. 73   R Development Core Team. 2009. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. Vienna, Austria. http://www.R-project.org  Raymond, M. and F. Rousset. 1995. GENEPOP (Version 1.2): Population Genetics Software for Exact Tests and Ecumenicism. Journal of Heredity 86:248-249.  Reichert, A.S. and W. Neupert. 2004. Mitochondriomics or what makes us breathe. TRENDS in Genetics 20(11):555-562.  Richards, J.G., B.A. Sardella, and P.M. Schulte. 2008. Regulation of pyruvate dehydrogenase in the common killifish, Fundulus heteroclitus, during hypoxia exposure. The American Journal of Physiology- Regulatory, Integrative and Comparative Physiology 295:979-990.  Ropson, I.J., D.C. Brown, and D.A. Powers. 1990. Biochemical Genetics of Fundulus heteroclitus (L.). VI. Geographical Variation in the Gene Frequencies of 15 Loci. Evolution 44(1):16-26.  Schulte, P.M. 2001. Environmental adaptations as windows on molecular evolution. Comparative Biochemistry and Physiology Part B 128:597-611.  Scott, G.R., J.T. Rogers, J.G. Richards, C.M. Wood, and P.M. Schulte. 2004. Intraspecific divergence of ionoregulatory physiology in the euryhaline teleost Fundulus heteroclitus: possible mechanisms of freshwater adaptation. The Journal of Experimental Biology 207:3399-3410.  Sejerkilde, M., J.G. Sørensen, and V. Loeschcke. 2003. Effects of cold- and heat hardening on thermal resistance in Drosophila melanogaster. Journal of Insect Physiology 49:719-726.  Skinner, M.A., S.C. Courtenay, W.R. Parker, and R.A. Curry. 2005. Site Fidelity of Mummichogs (Fundulus heteroclitus) in an Atlantic Canadian Estuary. Water Quality Research Journal of Canada 40(3):288-298.  Smith, K.J. and K.W. Able. 1994. Salt-Marsh Tide Pools as Winter Refuges for the Mummichog, Fundulus heteroclitus, in New Jersey. Estuaries 17(1B):226-234.  Somero, G.N. 2002. Thermal Physiology and Vertical Zonation of Intertidal Animals: Optima, Limits, and Costs of Living. Integrative and Comparative Biology 42(4):780-789.  Sweeney, J., L. Deegan, and R. Garritt. 1998. Population Size and Site Fidelity of Fundulus heteroclitus in a Macrotidal Salt Marsh Creek. Biological Bulletin 195(2):238-239.  Taylor, M.H. 1999. A Suite of Adaptations for Intertidal Spawning. American Zoologist 39(2):313-320.  Truchot, J.-P., and A. Duhamel-Jouve. Oxygen and Carbon Dioxide in the Marine Intertidal Environment: Diurnal and Tidal Changes in Rockpools. Respirin Physiology 39:241-254.  74  Van Oosterhout, C., W.F. Hutchinson, D.P. M. Wills, and P. Shipley. 2004. MICRO- CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4:535-538.  Wares, J.P. 2002. Community genetics in the Northwestern Atlantic intertidal. Molecular Ecology 11:1131-1144.  Watt, W.B. 1977. Adaptation at specific loci. I. Natural selection on phosphoglucose isomerase of Colias butterflies: biochemical and population aspects. Genetics 87:177-194.  Watt, W.B. 1983. Adaptation at specific loci. II. Demographic and biochemical elements in the maintenance of the Colias PGI  polymorphism. Genetics 103:691-724.  Watt, W.B., R.C. Cassin, and M.S. Swan. 1983. Adaptation at specific loci. III. Field behaviour and survivorship differences among Colias PGI genotypes are predictable from in vitro biochemistry. Genetics 103:725-739.  Weatherley, A.H. Effects of Superabundant Oxygen on Thermal Tolerance of Goldfish. Biological Bulletin 139(1):229-238.  Yeager, D.P. and G.R. Ultsch. 1989. Physiological Regulation and Conformation: A BASIC Program for the Determination of Critical Points. Physiological Zoology 62(4):888-907.  Zeyl C., B. Andreson, and E. Weninck. 2005. Nuclear-Mitochondrial Epistasis for Fitness in Saccharomyces cerevisiae. Evolution 59(4):910-914.   75  APPENDIX       Figure A1. Mean coefficient of variation between two trials measuring routine mass-specific metabolic rate in Northern (n=12) and Southern (n=8) subspecies of killifish acclimated to 20°C. There was no significant difference between groups (p=0.988) as assessed by One-way ANOVA. a a 76   Figure A2. Pearson correlation between the first and second mass-specific metabolic rate of the same fish from the Northern (n=12) and Southern  (n=8) subspecies acclimated to 20°C. (p=2.03x10-4) y= 0.6525x + 2.9476 R2=0.5449 p=0.000203 77                   Figure A3. Distribution of residuals of Mo2 compared to mitochondrial genotype (A) and compared to sample location (B). Residuals were presented as means in Figure 7. Red Bank Northern n=37; Metedeconk Northern n=11; Metedeconk Southern n=12. A B 78  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0069995/manifest

Comment

Related Items