Unlike most rodent models, prairie voles, show very few sex differences in social behavior. Despite the similarities in their behavior male and female voles, appear to use different physiological machinery to achieve the same behavioral goals. Therefore, male and female voles are likely to respond differently to pharmacological behavioral modifiers. To test whether the behavioral effects of serotonin potentiation vary by gender and/or reproductive context, I compared the effects of the selective serotonin reuptake inhibitor fluoxetine on parental and aggressive behavior in pairbonded, parentally-experienced male and female voles, and in pairbonded, parentally-inexperienced male and female voles. Because sociosexual experience influences the function of the serotonin system, the target of fluoxetine, I also compared serotonergic function in male and female voles that had mated, remained with a same-sex sibling, or been paired with a novel conspecific of the same sex. Fluoxetine increased the latency to parental behavior in parentally-experienced male and female voles and pairbonded, parentally-inexperienced male voles. Fluoxetine also decreased aggressive behavior in parentally-experienced male voles, but had no effect on the aggressive behavior of parentally-experienced female voles, or pairbonded, parentally-inexperienced voles of either sex. In addition, fluoxetine reduced serotonin turnover in the frontal cortex and hypothalamus/preoptic area of male and female voles. Serotonin turnover was also affected by sociosexual experience. Mating and cohabitation with a mate increased serotonin turnover in the hypothalamus of male and female voles. Furthermore, the levels of serotonin and 5-hydroxy-indoleacetic acid in the frontal cortex and amygdala were sexually dimorphic. Although this thesis focused primarily on the influence of gender and reproductive context on the behavioral effects of fluoxetine, it also revealed important aspects of parental and aggressive behavior that were independent of the effects of fluoxetine. Parentally-inexperienced female voles were infanticidal one week into pregnancy whereas their mates were parental. Furthermore, the aggressive behavior of parentally-inexperienced, pairbonded voles but not parentally-experienced voles was affected by the gender composition of the resident-intruder pair. Parentally-inexperienced, pairbonded voles were more aggressive with opponents of the same sex than opponents of the opposite-sex. The findings described in this thesis suggest that fluoxetine has sexually dimorphic effects on behavior and that sociosexual factors influence the effects of fluoxetine on behavior as well as the function of the serotonin innervation of the brain.

A large number of studies on both animals and humans have demonstrated that learning is influenced by context, or secondary cues that are present when an association is formed. The mechanistic aspects of context-dependent memory retrieval, or the context shift effect, have been well studied in strictly controlled laboratory settings. However, the adaptive value of attending to contextual cues in ecologically relevant situations has received considerably less attention. The goal of this dissertation is to examine the functional value of attending to context during learning. The jumping spider Phidippus princeps is capable of learning the ecologically relevant task of avoiding aposematic prey. However, when spiders were tested in an environment different from the one in which they were trained, attack rates increased and spiders no longer demonstrated retention of the association. Subsequent experiments demonstrated that spiders can use contextual cues to inform adaptive decisions about patch selection based on experience with prey of varying palatability or electric shock. Lastly, learning to avoid prey across a consistent contextual background versus a variable background produces initial differences in rates of learned avoidance, while ultimately learning performance between the groups is similar. Thus, contextual information (either in its constancy or variability) is an important component of avoidance learning in jumping spiders and can be valuable information in adaptive decision-making.

A rat model was used to determine whether prenatal cocaine exposure results in long-term changes in hippocampal-dependent contextual fear conditioning. Pregnant dams received either 40 mg/kg cocaine HCl SC (COC), an equal volume of 0.9% saline (SAL), or received no injections (UT) from gestational day 8 through 20. SAL animals were also pair-fed to COC subjects. Experiment 1 tested one-trial contextual fear conditioning in adult male offspring. Freezing and defecation were measures of fear. Prenatal cocaine exposure did not affect context conditioning, but there was an overall increase in SAL and COC defecation, indicating an increased generalized fear in these subjects. To better mimic binge cocaine use, COC dams in Experiment 2 and 3 received 20 mg/kg cocaine HCl SC, b.i.d. A preliminary open field task revealed that SAL offspring were more exploratory than UT controls and that females were more active than males. In Experiment 2, adult male and female offspring received 4 days of context conditioning and 3 days of no-shock extinction. During extinction, access to an adjacent chamber enabled the observation of four additional measures of fear: side crossing, latency, nose crossing, and side-differential. Experiment 2 repeated previous reports of gender-specific contextual fear. Males showed a greater level of freezing and defecation, higher latencies and side-differentials, and a lower level of side crossings and nose crossings than females. Prenatal cocaine exposure resulted in exaggerated gender-specific fear conditioning as measured by nose-crossing and side differential during extinction. Experiment 3 tested the effects of prenatal cocaine exposure on latent inhibition (LI) of contextual fear in year-old females. Vertical nose crossing (VNC), freezing, and defecation were measures of fear. LI was evidenced as an attenuation of freezing and VNC in pre-exposed (PE) animals compared to non-pre-exposed subjects. Prenatal cocaine exposure resulted in an enhanced LI effect. COC females showed a low level of baseline VNC, but COC-PE subjects showed a greater level of VNC than controls following the first shock during conditioning. The results of these experiments suggest that the effects of prenatal cocaine exposure on hippocampal-dependent learning are subtle, selective, and gender specific.