Hwei-yen Chen described her experiments to explain why aging (an increase with age in the chance of death per year) sometimes levels off at an "aging plateau." For example, humans are more likely to die in their 101st year than their 70th, but not more likely to die in their 101st than their 100th. Do we "get our second wind" and stop aging? Or is it just that older cohorts have already lost the frailer individuals (the heterogeneity hypothesis).
Chen exposed nematode populations to high vs. low death rates caused either by temperature stress (preferentially killing the weak) or random assassination. The aging plateau was only seen in the temperature-stress populations, consistent with effects from elimination of weaker individuals earlier in life.
I think I missed something, though. As I've described it, seems like this experiment could have been done with a single cohort of individuals, which would be increasingly enriched with stronger individuals as the weaker were killed off. But this was described as experimental evolution. If the evolution treatments preferentially killed weaker genotypes, wouldn't that eliminate the variation the heterogeneity hypothesis needs to work? Maybe someone else who saw this talk can clarify or maybe I can find her poster. Or maybe we'll have to wait for her paper.
Alan Cohen gave an interesting talk in the same session. He showed chance of death per year data as a function of age for a bunch of species. In humans and many other species, death rate increases with age. This has been explained by, for example, tradeoffs between longevity and early reproduction. But apparently many species don't have an increasing risk of death with age. So maybe such tradeoffs aren't universal?
In a theoretical talk, Olivier Cotto discussed the evolution of aging in metapopulations. Local extinction and recolonization processes were shown to favor evolution of different life-history strategies (reproductive effort as a function of age) for dispersing and nondispersing individuals.
Later, Will Ratcliff presented our latest data on the evolution of aging after experimental evolution of multicellularity, research led by Mike Travisano. Excellent experimental work by Jennifer Pentz has been key to our success. As discussed in our recent PNAS paper, we started with unicellular yeast and imposed selection for rapid settling in liquid media. Within about two weeks, most of our replicate populations evolved a "snowflake" phenotype: clusters of connected cells that settle much faster than their unicellular ancestor. Under continued settling selection, clusters competed with each other based on differences in growth, reproduction (via multicellular propagules), and survival (settling fast enough to make it into the 5% of the population that made it into the next generation).
The first snowflake clusters to evolve didn't age significantly faster than their unicellular ancestor. In other words, only a few of the cells within a cluster would die over the course of a few days, and the ability to compete with a reference genotype wasn't much less for older than younger clusters. But, under continued settling selection, faster aging evolved. Older clusters accumulated more dead cells and were less competitive than younger clusters.
Why did aging evolve? The dead cells form break-points that facilitate reproduction (release of multicellular offspring). And aging doesn't really set in until after 24 hours, when settling selection is imposed. Only 5% of cluster survive that selection, so there's only very weak selection against traits that only reduce fitness after 24 hours. Imagine if humans only lived 30 years. There wouldn't be much selection against genes that cause cancer in 80-year-olds.