Life Sciences Undergraduate Research Program
Symposium Abstracts:
Talk 1:
Population dynamics in a fragmented landscape with small patches: The Bodie pikas
Sabrina Jones1,2, Andrew Nemecek3, Lorida Llaci4, and John D. Nagy2,5
1School of Life Sciences, Arizona State University, 2School of Mathematical and Statistical Sciences, Arizona State University, 3Geography Department, University of Montana, 4Translational Genomics Research Institute, Phoenix, AZ, 5Department of Life Science, Scottsdale Community College
A population of American pikas (Ochotona princeps) inhabiting an anthropogenic landscape in the ghost mining town of Bodie, CA has historically been interpreted as a true metapopulation, where dispersal among patches of habitat plays a definitive role in its population dynamics. However, this assumption has never been explicitly demonstrated; in fact, it has been challenged by two competing hypotheses. The first suggests that, rather than patches being roughly equal in size and connectivity as in a metapopulation, large patches act as mainlands, making the landscape a classical MacArthur-Wilson island-mainland system. The second hypothesis suggests that observed occupancy patterns are a result of spatially correlated extinction events; in this hypothesis, dispersal plays a negligible role. Here we show, using 20 years of empirical patch occupancy data, that dispersal must be a key driver of the population dynamics of the Bodie pikas. Furthermore, a Hanski Incidence Function Model, which has become a standard modeling framework for metapopulations, fits the data better than do models of the other two hypotheses. In addition, the metapopulation concept has much more predictive and explanatory power. The Bodie pika population is well-suited to provide insight into fragmented population dynamics because it is distributed over discrete habitat patches, and we possess a series of high-quality censuses of the population from 1972 to 2010. It has become a standard empirical model of the effects of habitat fragmentation; therefore, it is critical that we have an accurate picture of the drivers of its population dynamics.
Talk 2:
Artificial selection for dispersal in experimental metapopulations of Tribolium confusum
Adam T. Hrabovsky1, Sarah H. Ung2, Michele V. Moreno2, Kerry J. Calhoun2,
Perry Olliver3, John D. Nagy2,4
1School of Life Sciences, Arizona State University; 2Department of Life Sciences, Scottsdale Community College; 3School of Molecular Sciences, Arizona State University; 4School of Mathematical and Statistical Sciences, Arizona State University
In metapopulations, dispersal connects subpopulations residing in discrete patches of habitat surrounded by uninhabitable matrix. In the 1970s Levins showed that metapopulation persistence requires that colonization rates equal extinction rates, which in turn requires adequate dispersal. Dispersal rate, on the other hand, is determined by evolutionary forces acting on individual fitness, not population persistence. The dynamics of this interplay are not entirely understood. Any experimental study of such dynamics requires a species in which dispersal has high heritability. Here we investigate the heritability of dispersal in artificial metapopulations of confused flour beetles (Tribolium confusum). We show that dispersal in T. confusum has a strong heritable component, but also exhibits a high degree of plasticity depending on environmental conditions. The key environmental determinant appears to be humidity. We also corroborate the results of Ogden and others who suggest that dispersal can be artificially selected in this species, which also supports the conclusion of high heritability of dispersal behavior.
Talk 3:
Evolution of treatment resistance in advanced prostate cancer
Khoa Dang Ho1, Paige N. Mitchell1, William J. Baker2, Jonathan Trautman2, Chandler M. Grant2, Alaina E. Daum2,3, and John D. Nagy2,4
1School of Life Sciences, Arizona State University; 2Department of Life Sciences, Scottsdale Community College; 3School of Molecular Sciences, Arizona State University; 4School of Mathematical and Statistical Sciences, Arizona State University
Recurrent and advanced prostate cancers are typically treated with total androgen blockade. However, androgen deprivation almost inevitably leads to castration resistance. Molecular mechanisms of castration resistance have been elucidated--the most common of which is amplification of the androgen receptor (AR) gene. But the ultimate cause of resistance remains unknown. Two hypotheses have been suggested: (i) resistance arises from cell plasticity; and (ii) resistance is caused by natural selection acting on mutant clones within the tumor. Here we show that evolution by natural selection is likely to be the ultimate cause of treatment resistance in prostate cancer. We found that, in a sample of 55 patients treated with intermittent androgen deprivation, the velocity of serum prostate specific antigen (PSA) decline tends to decrease over sequential on-treatment phases. In contrast, off-treatment PSA velocity exhibits no specific pattern in sequential cycles. These observations are consistent with treatment generating directional selection for castration resistance during treatment periods only. These results corroborate a predictive mathematical model that includes natural selection for AR expression under androgen deprivation. Such a model promises to be a key tool in managing castration protocols to mitigate the effects of treatment resistance in prostate cancer.