Why do animal populations boom and bust? This is the central question of population biology. Market squid (Doryteuthis opalescens) are an excellent study species for this question as they are fast growing, semelparous, and short-lived. Their rapid population growth allows them to support the largest fishery in California. However the population exhibits dramatic swings in abundance, periodically resulting in the complete collapse of the fishery. Despite their importance, relatively little is known about the drivers of their abundance. Here, I combine laboratory experiments, field observations, and novel statistical techniques to better understand what causes market squid to boom and bust. In the first half of my thesis, I perform a simulation study and find that nonparametric time series methods are particularly useful for understanding the dynamics of short-lived species such as market squid. Indeed, for short-lived, highly productive species, I find that parameter estimates of traditional mechanistic models may be severely biased when calibrated using conventional statistical techniques. Further, I perform a meta-analysis of the Ransom A. Meyers global stock-recruitment database and find that simple time series methods are able to uncover broad patterns of determinism in fish recruitment, thus validating their usefulness for understanding ecological dynamics. In the second half of my thesis, I combine these time series methods with laboratory and field observations to uncover the drivers of market squid abundance. In a mesocosm experiment, I obtain the first empirical estimates of the mortality rate and egg-deposition rate of spawning market squid. I also obtain unique insights into the life-history and vulnerability of spawning market squid. I then combine these insights with a study of market squid paralarvae collected in the 60-year CalCOFI program which includes a time series analysis, a growth analysis, and a size- distribution analysis. The time series analysis shows that variables acting on the juvenile and adult stage (but not the paralarval stage) are critical for forecasting future squid abundance. In agreement with this, the growth analysis shows that paralarvae grow quite well in the "bad" years of El Nino. Finally, the size-distribution analysis shows that paralarvae mortality is actually elevated during the "good'' years of La Nina, again indicating that the paralarvae stage is not the population bottleneck. This suggests, in contrast to conventional expectations, that the juvenile and adult stage are in fact the critical stages for determining population abundance. I conclude with a discussion of the research approach implemented here