Seminars

The Ecology and Evolutionary Biology Department has been working hard and monitoring, as well as planning, for COVID-19/Coronavirus and the safety of the UCLA community. We understand that we live in very uncertain times and, with news about the virus changing rapidly, it is hard to digest everything we are reading and hearing from all of the different news outlets. Please know the department is committed to working in everyone’s best interest—students, faculty, staff, and community at large.

We have been mandated by the Chancellor’s Office that there be no hosting of any in-person event/gathering/meeting, of any size, during the duration of Spring quarter 2020. With this, all departmental seminars are canceled for the quarter. We are looking to reschedule speakers in the upcoming academic year. We appreciate your understanding at this time.

March 12, 2020

5:00pm 1100 TLSB

CANCELLED-EcoEvoPub Series

" Graduate Student Presentations "

ALBERT CHUNG

Department of Ecology and Evolutionary Biology,

Campbell-Staton Lab, UCLA

 

“Sex-biased Gene Expression and Sexual Dimorphism in Anole Lizards”

 

Adult males and females of a species often possess differences in body size (sexual size dimorphism, SSD) despite the genomic constraint of a single, shared genome. SSD results from a variety of evolutionary pressures that result in the sexes possessing differing body size optima and represents a form of intersexual conflict. One genetic mechanism that may allow the sexes to overcome the constraint of a shared genome and achieve divergent body size optima is differential gene expression (i.e., sex-biased gene expression). To understand the role of gene expression in the evolution of SSD, we performed two RNA-seq experiments.

In experiment one, we compared brain, liver, and muscle gene expression between the sexes across ontogeny in a lizard species with extreme male-biased SSD (the Bahamian brown anole). We predicted that 1) sex-biased gene expression would increase during ontogeny and 2) these ontogenetic increases in sex-biased expression would differ between tissues. We found that sex-biased gene expression increased during development, but that the trajectory of this ontogenetic increase in sex-biased expression varied between tissues.

In experiment two, we compared brain, liver, and muscle gene expression between the brown anole and a species that is sexually monomorphic in body size (Panamanian slender anole). We predicted that brown anoles would 1) exhibit high levels of sex-biased expression of the entire transcriptomes of liver and muscle and 2) express growth regulatory networks dimorphically relative to the monomorphic slender anoles. We found that brown anoles do indeed exhibit higher levels of sex-biased expression of both entire transcriptomes and growth regulatory networks in the liver and muscle compared to slender anoles. By contrast, slender anoles exhibit higher sex-biased expression in the brain.

Ultimately, this work will increase our understanding of the gene expression mechanisms that may resolve intersexual conflict and facilitate the evolution of sexual dimorphism.

 

SHAWN SCHWARTZ

Department of Ecology and Evolutionary Biology,

Alfaro Lab, UCLA

 

High-throughput phenoscaping with deep learning for large-scale analyses of color pattern diversity

 

Deep learning, a branch of machine learning, can serve as a powerful toolkit for studies in ecology and evolutionary biology. In the era of big data, gaining more insight into complex patterns within a dataset can be transformative for conducting large-scale comparative analyses. One complex array observed in ecosystems is the immense diversity of color patterns. Coloration and patterning are ecologically important functions that can regulate optimal levels of conspicuousness, depending on the viewer, for attracting mates and avoiding predators. These proposed functions of color are important to study within a comparative framework to better understand the evolutionary history of these immensely diverse traits; however, objectively quantifying color pattern has been a challenge that has limited efforts to perform large-scale analyses of these color patterns. Additionally, preparing big datasets necessary for quantifying color pattern complexity has previously taken substantial time and effort that have contributed to a data bottleneck for these types of studies. Here, we present a quantitative pipeline for performing high-throughput phenoscaping on color patterns using images of digitized specimens of coral reef fishes. Given that popular pre-trained models for machine learning algorithms are inherently anthropocentrically and terrestrially biased, we provide an updated, robust model that accounts for these biases and corrects them to then perform automatic image processing and color quantification for fishes at an unprecedented rate.