Gulf Killifish Fundulus grandis is an estuarine species used as a live bait for marine sport fishes, native from Veracruz, Mexico, along the Northern Gulf of Mexico coast, to the eastern coast of Florida. Culture protocols are established, but Gulf Killifish have not been fully adopted as a commercially produced species by producers, possibly due to economic efficiency and needs for advances related to inland production. Therefore, production methods from spawning in recirculating aquaculture systems (RAS) to grow-out to market size in earthen ponds were examined. Increasing broodstock density or spawning substrate surface area did not increase egg production in RAS. In freshwater, juvenile Gulf Killifish exhibited broad pH and hardness tolerances, and multiple cohorts were stocked and grown to market size in one season in earthen ponds, with rapid growth rates at 50,000 fish/ha. After low-salinity rearing, osmotic stress differed between cohorts, but overall salinity tolerance was not affected.

The turquoise killifish Nothobranchius furzeri is an increasingly popular model species for comparative vertebrate research. As such, the basic physiology, including responses to one or multiple stress events are of primary interest to the present study. After successful rearing, adult male and female killifish were exposed to one or more acute confinement events. Whole-blood was taken from adult males, in addition to male and female whole-body samples for cortisol analysis. Separate adults were also sampled for tissue specific expression of corticotropin releasing hormone (CRH), mineralocorticoid receptor (MR), and glucocorticoid receptor (GR) mRNA within the cranium, gills, and liver. Following a 30 minute confinement stress, male plasma cortisol significantly differed from baseline at 30 minutes (P=0.04). Similarly, both male and female whole-body cortisol were significantly increased (P=0.004 and P=0.04, respectively) at 15 and 30 minutes post-stress. Whole-body cortisol did not differ between the sexes at any sampling point, however; expression of gill MR at 15 minutes was significantly higher in males (P=0.05). Exposure to daily repeated confinement affected the cortisol response in both males and females, resulting in lower baseline and 60 minutes post-stress values in repeatedly-stressed males (P=0.04 and P=0.006, respectively) and lower cortisol at 30 and 120 minutes post-stress in repeatedly-stressed females (P=0.04 and P=0.04). Repeated exposure also resulted in increased cranial CRH and MR at 15 minutes post-stress (P=0.02 and P=0.05, respectively) compared to singly-stressed males. In females, repeated exposure increased gill MR at 120 minutes (P=0.05), but a single stressor resulted in relatively greater expression of cranial CRH at 120 minutes (P=0.02) and MR at 15 minutes (P=0.05). Collectively, the reduced cortisol production observed in repeatedly-stressed adults coupled with only transient changes in receptor expression suggest acclimation to daily stressors can be detected in as little as one week in adult turquoise killifish. This is also the first description of the stress response on this important model species.

Largely through the use ofin vitro preparations there is now a generally accepted theory for ion transport for the seawater (SW) gill. However, to date there is no generally accepted freshwater (FW) model for the mechanisms of NaCl transport in the teleost gill. By using an Ussing chamber approach with the opercular epithelia of Fundulus heteroclitus and Oreochromis niloticus, and the urinary bladder of Oncorhynchus mykiss, all acclimated to FW, we hoped to establish one as a possible model for the study of FW ion regulation. FW Fundulus opercular epithelia displayed a serosal negative transepithelial potential (Vₜ) of-43.9 mV, transepithelial conductance (Gₜ) of 1.94 mS·cm⁻², and active transport of Cl⁻ from the mucosal FW against a strong electrochemical gradient. Na⁺ movement was dominated by passive diffusion. The opercular epithelia of Fundulus adapted to 10% SW exhibited properties similar to SW Fundulus by actively extruding Cl⁻ while Na⁺ moved passively into the mucosal 10% SW. With FW bathing the mucosal surface, FW Oreochromis opercular epithelia displayed a serosal positive Vₜ of +8.0 mV, Gₜ of 1.78 mS·cm⁻², and active reabsorption of Na⁺, Cl⁻ and Ca²⁺ against large electrical and/or chemical gradients. The FW Oreochromis opercular epithelia is the only FW in vitro preparation to date that exhibits active absorption (albeit at small absolute rates) of both Na⁺ and Cl⁻. FW 0. mykiss urinary bladders mounted in vitro under symmetrical saline conditions exhibited a transepithelial conductance (Gₜ) of-9.15 mS·cm⁻² and electroneutral active absorption of Na⁺ and Cl⁻ from the mucosal urine side. The transport of Na⁺ and Cl⁻ was a partially coupled process whereby removal of Na⁺ from the mucosal saline decreased Cl⁻ absorption by a 56% and removal of Cl⁻ inhibited Na⁺ absorption by 69%. However, active net absorption of both ions persisted when the counter-ion was replaced with a non-permeant ion. Under more realistic conditions with artificial urine bathing the mucosal surface, Vₜ increased to a serosal positive ~+7.6 mV and Gₜ decreased to ~1.47 mS·cm⁻² Unidirectional influx rates of both Na⁺ and Cl⁻ were much lower, but active absorption of both ions still occurred. Replacement of Na⁺ in the mucosal artificial urine caused no change in unidirectional influx of Cl⁻ and vice versa. The mucosal addition of DIDS, amiloride or bumetanide (10⁻⁴M) all had no affect on absorption rates of Na⁺ and/or Cl⁻, under either artificial urine or symmetrical saline conditions. When the mucosal surface was bathed in artificial urine, removal of mucosal Cl⁻· significantly reduced the maximum transport rate (Jₘₐₓ of Na⁺ (6.1 ~2.1 μmol·cm⁻²·h⁻¹) but had no effect on affinity for Na⁺ (Kₘ~27 mM). Similarly, removal of mucosal Na⁺ significantly reduced the Jₘₐₓ for Cl⁻ uptake (11.4 ~2.4 μmol·cm⁻²·h⁻¹) but had no effect on Cl⁻ Kₘ(~37 mM). The anterior portion of the urinary bladder transported Na⁺ and Cl⁻ at a faster rate than the posterior portion under symmetrical saline conditions, but there was no difference in measured Na⁺/K⁺ -ATPase activities between the two portions. In conclusion, opercular epithelia of Fundulus and Oreochromis demonstrated transport properties believed to be typical of FW teleosts; either preparation may eventually prove to be a good working model for studying FW transport mechanisms. The urinary bladder of Oncorhynchus mounted in vitro did not behave characteristically as the teleost gill is thought to function, but this model may be useful in characterizing various other mechanisms of ionic transport. / Thesis / Master of Science (MSc)

Chronically contaminated killifish from Newark Bay (NB) NJ, and New Bedford Harbor (NBH) MA, have developed resistance to halogenated aromatic hydrocarbons that bind to and activate the aryl hydrocarbon receptor (AHR). To study the mechanisms of resistance, adult killifish were exposed to halogenated and non-halogenated AHR ligands and enzymatic and toxicological endpoints were measured in adult and embryonic fish. The chlorinated and non-chlorinated AHR ligands 3,34,4-tetrachlorobiphenyl (PCB77) and benzo-a-pyrene (B[a]P) induced cytochrome P450 1A (CYP1A) in reference site, but not in NB killifish. Expression of CYP3A (not part of the AHR gene battery) was inducible only in Flax Pond killifish. Basal expression of the phase II enzyme glutathione-s-transferase (GST) was higher in NB killifish. These results suggest that NB killifish are resistant to CYP1A induction by chlorinated and non-chlorinated AHR ligands. Higher basal GST activity observed in NB killifish could be protective against toxicity caused by contaminants found in this site. Activation of AHR and induction of CYP1A, by AHR ligands has been associated with the toxic effects caused by these chemicals. To determine the association between resistance to CYP1A induction and the toxicity caused by AHR ligands, CYP1A activity, developmental deformities and reactive oxygen species (ROS) production were measured in reference site and contaminated (NB and NBH) killifish embryos exposed to AHR ligands. 3,34,45-pentachlorobiphenyl (PCB126) and 3-methylcholantherene (3-MC) induced CYP1A, and ROS production in reference site embryos. NB and NBH embryos were resistant to PCB126 induction of CYP1A, but responded to 3-MC. Killifish embryos from NB and NBH were resistant to PCB126 induced deformities. PCB126 and 3-MC did not increase ROS production in NB or NBH killifish embryos. Alpha-naphthoflavone (ANF) (an AHR/CYP1A inhibitor) blocked PCB126 mediated deformities and CYP1A induction in reference site embryos, but increased ROS production. The P450 inhibitor, piperonyl butoxide (PBO) was able to block PCB126 mediated induction of CYP1A activity and ROS production. These results suggest that PCB126 induced deformities are dependent on activation of AHR and CYP1A induction. In chronically contaminated killifish populations, loss of sensitivity to coplanar PCBs and PAHs could be through reduced expression of AHR, or altered DNA sequence or methylation status of the CYP1A gene promoter. Hepatic AHR expression, measured by photoaffinity labeling, was lower in NB killifish than reference site animals, suggesting that NB killifish express less AHR protein. DNA sequence analysis did not reveal considerable differences between contaminated and reference site populations, however additional DNA fragments were observed in some promoters but not in others. The methylation of the CYP1A promoters was studied using methylation sensitive restriction enzymes and no differences were detected between reference site and NB killifish. Treatment with the DNA methyltransferase inhibitor AzaC did not restore CYP1A induction by PCB126 in NB killifish. These studies suggest that resistance to activation of AHR and induction of xenobiotic activating enzymes (CYP1A and CYP3A) in combination with increased expression of conjugating enzymes (GST) protects chronically contaminated killifish against these chemicals.

Accumulating evidence from epidemiological, wildlife, and laboratory studies indicates that abnormalities of reproduction, development and physiology can be ascribed to environmental contaminants with biological activities. Many such contaminants disrupt essential hormone-regulated processes by virtue of their ability to interact with nuclear hormone receptors (endocrine disrupting chemicals, EDC). Of particular concern are chemicals that mimic or block estrogen signaling (xenoestrogens, XE) through their direct interaction with estrogen receptors (ER). The current model of XE action focuses on disrupted transcriptional activity, as measured by changes in the expression of ER-regulated genes. However, transcription is tightly coupled to splicing, by which a single target gene transcript is processed to multiple structurally and functionally different mRNAs. In theory, any XE that interacts with ER to regulate transcription has the potential to disrupt splicing, thereby affecting not only mRNA quantity but also quality. To address this hypothesis, alternative splicing of the gene encoding ER alpha (esr1), itself an estrogen responsive gene, was investigated. In these studies, killifish (Fundulus heteroclitus), an environmentally relevant species, and zebrafish (Danio rerio), an advantageous laboratory fish model, were used. First, the occurrence of ER alpha splice variants in adult tissues, in developing embryos and in response to estrogens in the two species was documented. Additionally, the effects of long-term, multigenerational XE exposure on ER alpha splicing were examined in two killifish populations, one from an estrogenic (polluted) site and a second population from a reference (unpolluted) site. A subset of ER alpha variants from killifish was expressed in cell culture to document their transcriptional activities. To determine the in vivo relationship between estrogen responsiveness and an ER alpha splice variant of interest, esr1 splicing was experimentally altered in living embryos by microinjecting morpholino oligonucleotides, and changes in induction of a panel of estrogen responsive gene targets were measured as markers of effect. These results provide evidence that dysregulation of mRNA processing is also a mechanism of XE action, and suggest that resultant ER alpha splice variants mediate the short-term effects of estrogen disruption and are also part of the adaptive response to long-term, multigenerational XE exposures in the natural environment.

It is widely documented that non-human organisms express individual differences in behavioural patterns. For example individuals can be categorised as bold or shy and when these individual behavioural differences are consistent through time, they are termed behavioural types (BTs). In recent years research has identified that BTs often correlate across contexts/situations and these correlations are referred to as behavioural syndromes. Behavioural types and syndromes (i.e. personality) have also been implicated as major factors shaping population dynamics and the ability to buffer environmental disturbance. Recent theoretical predictions have proposed that BT variation may be underpinned by life-history strategies; however, these predictions have been little studied to date. Moreover, little research has focused upon environmental influences and the ontogeny of personality. In this thesis I use the Mangrove killifish (Kryptolebias marmoratus), a naturally occurring clonal vertebrate, as a model organism. This species presents a powerful tool providing the ability to replicate within and between isogenic genotypes in a controlled manner. Moreover the natural clonality expressed by this species permits environmental effects upon BT plasticity and BT-life-history interactions to be investigated within a developmental framework. In chapter 2, I present microsatellite genotyping results which show that the founding individuals used to propagate a laboratory population at The University of Exeter represent 20 genetically distinct homozygous genotypes. I additionally address five research questions exploring genotypic, environmental, and developmental effects upon three commonly studied BTs (exploration, boldness and aggression): Firstly; I ask do adult hermaphrodite and secondary males exhibit personality i.e. repeatable BT expression? In chapter 3, I present results showing that both of the sexes express short term personality. Moreover, I show that that genotype is an important factor influencing BTs expressed, regardless of sex, indicating underlying genetic control. Secondly I ask; does genotype level life-history variation underpin personality trait variation during ontogeny? In chapter 4, I show considerable developmental plasticity in behavioural expression between genotypes but not life-history and I find limited behaviour-life-history relationships during development. Thirdly I ask; does the rearing environment influence life-history and behavioural plasticity? In chapter 5, I show that in comparison to a control treatment, the presence of conspecifics during ontogeny results in an average reduction in behavioural scores; however, life-history was unaffected. In addition, I show that development in a low food environment lowered average exploration and growth rate but had no effect on boldness or aggression. Furthermore, fish exposed to a predation risk simulation during ontogeny exhibited similar behavioural scores as the control, yet this treatment generated BTs i.e. personality. My fourth question asks; does the parental rearing environment (utilised in chapter 5) influence behavioural expression in the next generation? In chapter 6, I show that transgeneratonal effects of each parental rearing environment influenced life-history but had a minimal effect upon behaviour in the next generation. Finally I ask; does kin or familiarity influence plasticity in associations and aggression? In chapter 7, I show that genotypes have the ability to discriminate kin and familiars and modulate aggression and association accordingly. These results support the concept that developmental and environmental induced plasticity may be more important than life-history in shaping behaviour. Furthermore, although adults exhibit personality and genotypic effects appear important, genotype interacts with environmental/experiential influences to differentially shape behavioural plasticity during ontogeny. I suggest that theoretical predictions regarding life-history may be insufficient to explain the complexity of animal personality in this species. I discuss these results within developmental and epigenetic frameworks with reference to the ecological significance of these patterns within this species and the animal kingdom as a whole.

The mangrove killifish, Kryptolebias marmoratus, has the potential of becoming a strong model organism for a range of biological disciplines thanks to its ability to self-fertilise, a process only known to occur in invertebrate animals until its discovery. Selfing, a natural occurrence in this species, has lead to the formation of clonal lineages composed of highly homozygous individuals. The aim of this thesis was to further establish K. marmoratus in the field of developmental biology by providing an information infrastructure to help advance research on this peculiar animal and further promote its place in the pantheon of model organisms. To do so, I first set out to standardise K. marmoratus embryology by providing defined developmental stages with clear visual representations of key embryonic structures. This staging series is an essential tool that will ensure repeatability and consistency within and across different laboratories. Secondly, I examined several techniques for embryonic manipulation and for imaging that can be used in an array of experimental designs. Using these techniques I demonstrated microinjection of embryos by monitoring the yolk syncytial layer and its nuclei, and time-lapse analyses of the yolk surface during embryonic development. Finally, I applied the knowledge gained from my first two studies and examined Bmp signalling in K. marmoratus embryos and its influence on body patterning. By inhibiting this pathway, I found a new phenotype characterised by an extremely short and split body axis. These data highlighted the importance of studying known signalling pathways in unknown organisms as species-specific differences may improve our understanding of fundamental developmental processes. This thesis demonstrates that with its easily obtainable and manipulated embryos, K. marmoratus can be used for embryological research in the same light as other model organisms such as zebrafish or medaka. The rising amount of information on mangrove killifish will help further take advantage of this unique and intriguing species, and supports the use of this hermaphroditic vertebrate as a strong comparative model in developmental biology.

Austrofundulus limnaeus is a species of annual killifish which inhabits ephemeral ponds in South America. The species is able to survive seasonally desiccating ponds due to their ability to produce robust embryos. The embryos of this species are capable of entering a developmental arrest, termed diapause II, which precedes the onset of drought. While in this arrested state embryos exhibit the greatest tolerance to anoxia of any characterized vertebrate at 25ºC. Furthermore, when raised at 30ºC, embryos escape the entrance to diapause II and go on to develop directly. Currently, little is known about the molecular mechanisms which induce and maintain this developmentally arrested state. In this study I have developed methods to analyze changes in histone modifications in the context of diapause II. Histone modifications were chosen due to their extreme conservation and well characterized role as modulators of gene expression in other systems. Results utilizing adapted immunobased assays show significant changes in the global amount of H3S10P, H3K27me and H3K4me, as the embryos progress from early embryogenesis through the exit of diapause. Additionally, it is revealed that there exists a degree of phenotypic plasticity with regards to the entrance into diapause II which is modulated by the environment (severe hypoxia 0.1% O2). This work builds a foundation for future histone modification studies and contributes the development of several tools to the field. This study contributes to a greater molecular understanding of the cue(s) which influence the remarkable phenomenon of obligate developmental arrest in a vertebrate embryo.

This research examined the effects of acclimation to lowered salinity, elevated temperature, and hypoxia on aerobic metabolism of the Gulf killifish, Fundulus grandis, a common estuarine resident of the Gulf of Mexico. Standard metabolic rate (SMR), maximum metabolic rate (MMR), absolute aerobic scope (AAS), and critical oxygen tension (Pcrit) were each influenced by one or more acclimation treatments. Assessing the consistency of these traits measured in the same individuals over time, all were found to be significantly repeatable with no indication that the repeatability of any traits was affected by acclimation conditions. Significant correlations were found between SMR and Pcrit (positively correlated), between SMR and AAS (negatively correlated), between MMR and AAS (positive), and between AAS and Pcrit (negative). This study, therefore, documents the effects of acclimation on these traits, their repeatability, and correlations among them. It further suggests that repeatability of these traits is not context dependent.

Modeling is a powerful tool that can be used to identify important relationships between organisms and their habitat (Guisan & Zimmermann, 2000). Understanding the dynamics of how the two relate to one another is important for conserving and managing ecosystems, but the extreme complexity of those ecosystems makes it very difficult to fully diagram. Unlike many other modeling techniques, Multivariate Regression Trees (MRTs) are not limited by a prior assumptions, pre-determined relationships, transformations, or correlations. MRTs have the power to provide both explanation and prediction of ecological data by producing simple models that are easy to interpret. This study proposed to use MRTs to evaluate and model relationships between Lucania parva and the environment and habitat of Florida Bay. Counts were transformed to presence-absence and abundance groupings. Models were first run using a variety of combination of response variables and all explanatory variables. Results of these models were used to select the best combination of response and explanatory variables in an effort to create a best fit model. Models indicated that Lucania parva populations are found in the dense (cover ≥50%), shallow water (<1.8 m) grass beds that occur in the western portion of Florida Bay. A best fit model was able to explain 63.7% of the variance with predictive error of 0.43.