Fall 2015 schedule
Note: Priority is given to graduate students. A ⊛ symbol next to the speaker's name means that approval is pending for a week and graduate students can still claim the slot.
Titles and abstracts
-
Lianas Affect Population Growth Rates of Tropical Trees Marco Visser Lianas are increasing across the Neotropics and dramatically reduce the growth, survival, and fecundity of their host trees. Whether lianas affect tree species population growth rates differentially remains unknown. In this talk, I show how we used integral projection models to integrate liana effects across the life-cycle into total effects on population growth rates for 33 tropical tree species from Panama. Liana infestation decreased tree growth, survival and reproduction. Fast-growing tree species exhibited stronger declines in survival than did slow-growing species. Heavy liana infestation involving more than half of the host’s canopy reduced population growth rates by 1.9% per year. The reduction was greatest among fast-growing species, which showed much higher sensitivity to liana infestation. Lianas have strong differential effects on tree species populations, and thus ongoing increases in liana abundance are likely to alter tree species composition in Neotropical forests.
Back to schedule -
How to reverse-engineer your social system in 7 easy steps Matthieu Barbier Societies are among the most salient examples of adapted systems with a hierarchy of goals - in biological terms, multiple selection levels operating simultaneously - from individual preferences to institutional efficiency and so forth. Much theoretical effort has focused on optimal or evolutionary responses to a single level of selection, or simple conflicts between the smallest level (individual wants) and the largest (collective good). Far less has been done to understand how and when multiple intermediate levels can develop and interact. I suggest that the clearest language for this problem is not agent or population dynamics, but a non-dynamical form of system modelling that looks more like (reverse) engineering. Interestingly, this approach is also far more compatible with the questions and data of qualitative social science, and may offer a way of making those more easily testable. What it lacks is a consistent formal framework applicable to natural systems, which I have tried to develop in a restricted setting.
Back to schedule -
Costly punishment as a strategy for optimising public goods investment Joanna Bryson [With the goal of] account[ing] for the variation by global region in antisocial punishment (ASP-sanctioning those contributing more than the sanctioner in public goods games (as described by Herrmann, Thöni & Gächter (2008)), [we] examined strategies expressed by Herrmann et al's subjects. We also looked at the impact of having a group member invest all (Cooperator) or none (Freerider) of their allotment in the PG. We created evolutionary ABM exploring the coherence of our hypotheses. In Herrmann et al's data, rates of ASP vary more than those of altruistic punishment (AP). Variation across subject pools in the proportion of Cooperators is much greater than in the proportion of Defectors. We found no significant correlation between AP and the proportion of Cooperators, AP and Freeriders, and ASP and Freeriders, but a strong anticorrelation between ASP and the proportion of Cooperators (r = −0.62, p < 0.01). Many more individuals employed both AP and ASP than ASP exclusively. Exclusive AP and complete abstention from punishment were widely-used strategies. When an individual is punished, they are less likely to invest the same amount in the PG in the next round. ASP results in random variation, whereas AP results in increased public goods investment (PGI). Our models show that when PGI is subject to selection in regularly varying contexts, populations form with peaks of similar PGI strategies across a range of values extending both above and below optimal PGI levels. We believe that overall levels of PGI are generated by populations composed of multiple strategies, including over and under investment (cf. Maclean et al 2010). Shifting between strategies as a result of individual experience may allow agile tracking of changing economic landscapes. Punishment may be a social mechanism accelerating such tracking. We are currently studying its impact with further simulations.
-
The Moran Process and Evolution in Finite Populations Daniel Cooney This talk will provide an introduction to the frequency-dependent Moran process as a model for stochastic evolution in finite populations. In particular, we will focus on the relationship between payoffs in underlying evolutionary games and the mapping between individuals' payoffs and their reproductive fitnesses in the Moran process. We will discuss the traditional linear fitness version of the Moran process, as well as the newer exponential fitness mapping and it's particularly nice properties. After outlining the standard tools for analysis of the Moran process, we will make use of these tools to explore the roles of assortment and kin selection in the evolution of cooperation.
Back to schedule -
Social-ecological dynamics in exploited marine systems Anieke van Leeuwen Over-harvesting of marine resources has resulted in a multitude of fish population collapses in recent decades. In order to better understand the processes underlying these collapses and to assess the potential for population recovery, fisheries management is moving away from both single-species stock assessments and total-allowable catch quota. Management approaches increasingly focus on the intricate dynamics in entire ecosystems as well as on the social dynamics of the fisheries. I will present an approach to couple ecological dynamics in simplified food web modules of exploited fish species with the social dynamics in the fisheries harvesting from the food web.
Back to schedule -
Testing for Directional Change in Carbon Budgets of Tropical Forests Using Forest Plot Data Helene Muller-Landau Tropical forest structure, dynamics, and tree functional composition vary (and covary) greatly within and among sites, with associated variation in carbon budgets, plant diversity, etc. Historically, most analyses of these characteristics have relied on ground-based censuses of tree trunk locations and diameters, though these have obvious limitations. In the last year, my lab has been conducting a pilot study using camera-carrying unmanned aerial vehicles (aka drones) and structure-from-motion software to map the structure of the top of the forest canopy in three dimensions and at high spatial (10 cm) and temporal (2 week) resolution. I will present preliminary results on forest structure, canopy gap dynamics, liana coverage, leafing and reproductive phenology of canopy trees, and characterization of tree species life history strategies based on these datasets. I will also describe some ongoing challenges and seek input on a grant proposal to extend this work to additional sites and time periods.
Back to schedule -
Strategic MABs and Social Foraging Models Karla Kvaternik The standard multi-armed bandit problem is a canonical prototype model of decision making under uncertainty. Classically, the problem involves a single decision maker faced with a finite set of options, each of which yields a random reward drawn from an unknown distribution each time it is sampled. The decision maker's task is to select a sequence of options in a way that maximizes her cumulative expected reward. This problem has provided interesting insights in a variety of applications in engineering, economics, social sciences and ecology. In this talk, I will consider several extensions of the classical MAB setting and explore their potential applications to the development of optimal foraging models for patchy resource landscapes. In particular, I will consider a multiplayer variant in which the sampling policies of some players may affect the rewards collected by other players. In the context of animal foraging, an "arm" corresponds to a resource patch, and the reward realized by sampling this arm may represent the efficiency with which an animal is able to consume the resource. Thus, the strategic MAB setting reflects the possibility that this efficiency is affected by the presence or absence of other agents at the same feeding patch. I will also draw analogies between this strategic MAB setting and the short-term decision making tasks faced by fishers in commercial fishery settings.
Back to schedule -
Changing Climate and Evolving Marine Ecosystems Colin T. Kremer Ecosystem responses to climate change depend on the collective reactions of many species, which individually can evolve, migrate, or decline. Evolutionary responses will be particularly important in areas that experience extreme or novel conditions, such as the tropics. Ecosystem function in these regions will depend on the pace of climate change, the rate at which species can adapt, and the consequences of maladaptation. Ecosystem models, embedded within sophisticated physical/chemical models of the earth, are powerful tools for exploring climate change scenarios. However, in an effort to simplify biological complexity, they typically make inconsistent assumptions about evolution or ignore its effects entirely. I will present preliminary results from a global marine ecosystem model that highlight how including evolutionary responses of plankton to rising ocean temperatures can affect marine ecosystems. I will show how the nature of these effects depends on trophic and competitive interactions between functional groups that may have very different adaptive capacities. Finally, using this project as a case study, I may give in to temptation and speculate about how we might efficiently and parsimoniously account for evolutionary dynamics in other ecosystem models.
Back to schedule -
Simple predictions of forest structure from resource limitations and allometric scaling Chris Kempes One of the main objectives in ecology and the Earth sciences is to uncover small sets of parameters that predict the dominant behavior of a biological system. For vascular plants it has been observed that scaling laws (also known as power laws) describe trends in many organism traits based primarily on organism size. Yet there remain many questions regarding the utility of these power-laws, how and why individual species deviate from central trends, and why variation exists in the overall allometries. I have developed a theory that synthesizes scaling laws, the morphology of trees, and resource limitations to predict the biogeography of maximum tree height. In this talk I will show how the characteristic of maximum tree height indicates forest structure along with a wide variety of other fundamental forest features. Furthermore, I will discuss how this theory generates simple predictions for the response of forest features to shifts in environmental conditions, including precipitation and temperature. Finally, I will discuss how this approach can be used to interpret and incorporate allometric shifts.
Back to schedule -
Linking dispersal, mortality and herbivory on coral reefs Lisa McManus Connectivity through larval dispersal is thought to be a major determinant of coral persistence, especially in terms of the recovery potential for populations that suffer catastrophic mass bleaching events. However, our understanding of the interactions between coral recruitment and mortality on reefs is limited, particularly when reproduction occurs in a discrete, seasonal manner as in the majority of stony coral species in the Indo-Pacific. Here, I will present a spatially implicit coral-algal competition model that incorporates coral larval recruitment under scenarios in which recruits are contributed both seasonally and continuously in time. In addition, I will address the controversial idea of bistability between coral- and algal-dominated regimes on reefs by taking a closer look at the dependency of herbivore grazing rate (i.e. algal mortality) on coral cover.
Back to schedule -
Social and Environmental Influences on Movement in Wild Baboons Ariana Strandberg-Peshkin When deciding where to move, social animals must integrate information about the habitat around them with social factors such as the locations and movements of conspecifics. Yet differentiating the role of social and physical environmental factors in shaping individual movement decisions remains a challenge. Here I will present recent work that combines simultaneous high-resolution GPS tracking data from almost an entire troop of wild baboons with a detailed three-dimensional reconstruction of the habitat to investigate the interplay between social and non-social factors in influencing the fine-scale movement decisions of baboons. I am using a step selection framework to quantify the features that are most predictive of where individual baboons choose to move, and to allow visualization of the jointly social and environmental “preference landscape” that underlies movement decisions. This is an ongoing project - feedback and ideas for how to extend the analysis will be much appreciated!
Back to schedule -
Future global finfish aquaculture under climate change Dane Klinger Most aquaculture is currently land-based, but future production will likely expand into the ocean due to increasing competition for inputs to on-land and coastal aquaculture and growing demand for seafood. The potential for offshore aquaculture to increase seafood production remains uncertain. We employ a novel spatial model to estimate the production potential of offshore finfish aquaculture given physical, biological, and technological constraints. Production potential is compared across species and regions and with climate change, based on outputs of a recent high-resolution global climate model. Implications for management and regulation of the emerging offshore aquaculture industry are discussed.
Back to schedule -
Social-ecological complexity in common pool resource use: Social foraging and management action in fisheries Emily Klein Insight into human behavior and decision making within the context of complex social and ecological systems is critical for understanding and managing resource use. I will present findings from our theoretical agent-based model of fishermen behavior, which provided insight into diverse incentives and their interactions with target species ecology and management action. Results demonstrate differences in fishermen behavior both between fishing sectors, but also among individuals within sectors, and potential consequences for managing resource use. These show that, due to individual experiences, behavior will differ both within and across groups, and may have strong implication for how people respond to management actions. Finally, our work reveals there can be a range of potential driving motivations behind modeled human behavior, with implications for how we address and understand human interactions with common pool resources.
Back to schedule -
Title George Hagstrom Abstract
Back to schedule
Links to previous schedules
- Fall 2000
- Spring 2001
- Fall 2001
- Spring 2002
- Fall 2002
- Spring 2003
- Fall 2003
- Spring 2004
- Fall 2004
- Spring 2005
- Fall 2005
- Spring 2007
- Fall 2007
- Spring 2008
- Fall 2008
- Spring 2009
- Fall 2009
- Spring 2010
- Fall 2010
- Spring 2011
- Fall 2011
- Spring 2012
- Fall 2012
- Spring 2013
- Fall 2013
- Spring 2014
- Fall 2014
- Spring 2015