In sexually reproducing organisms, Hamilton.s forces of natural selection lead to three distinct life stages: development, aging, and late life. Recently a fourth life stage known as the "death spiral" has been documented in experiments using Drosophila. The fecundity death spiral refers to a significant decline in the fecundity of females that are within 6-15 days of death compared to same-aged females that are not near death. This dissertation focuses on physiological changes that occur in aging, late life, and the death spiral.

Chapter 1 reviews theory and experiments for explaining late life mortality rate plateaus. Chapter 2 tests populations of Drosophila for physiological transitions between aging and late life. Different physiological characteristics that deteriorate during aging vary with respect to their continued deterioration, acceleration, or deceleration in late life. It appears that late-life physiology is complex and is distinct from the physiology of aging.

Chapter 3 shows that male virility also plateaus in late life, as is predicted from the evolutionary theories for late life mortality plateaus. In addition, the results also show that, like females, males experience a reproductive death spiral; the virility of males within seven days of death is significantly lower than the virility of same-aged males that are not near death.

Three-day female fecundity data have in the past been shown to be predictive of individual flies' time of death. Chapter 4 uses such predictions to dichotomize a population of same-aged female flies into "death spiral" and "non-spiral" groups. Females that are in the death spiral have significantly more deteriorated physiology compared to same-aged females that are not near death.

Chapter 5 tests for physiological abnormalities in heterozygous dOpa1^+/- mutants and shows that heterozygous dOpa^+/- is associated with multiple organ abnormalities, including heart dysfunction, in an age-dependent manner. Chapter 6 compares heart phenotypes in experimentally-evolved populations of Drosophila. There are significant differences in heart function among these populations in the absence of large-effect mutations, suggesting that heart dysfunction in these populations results from many genes and many pathways with small effect.