Titles and abstracts
Wednesday September 19th at 12:30pm
Robustness of cooperation under variable social and environmental conditions
The sustainable use of local and global commons is dependent on cooperation amongst its users that restrain themselves from selfish profit maximization that ultimately leads to overexploitation. Empirical evidence has shown many positive examples of successful cooperation despite grim theoretical predictions that claim it impossible. Rules and norms that have emerged with those communities have been identified as crucial for success. However, with global change the resource and social conditions under which cooperation is achieved may change dramatically. Climate change for example will affect water flows making them more variable and uncertain which increases the pressure on cooperation. In this contribution we want to assess the robustness of cooperation to variable resource conditions in a community where cooperation is maintained through a norm that ostracizes members that are benefiting more than the average. We use an evolutionary agent-based model to identify cooperative outcomes under different degrees of resource variability. The results show that low variability favors defectors while high variability benefits co-operators. This corresponds well with recent empirical results that suggest that supply variability in international bilateral river basins creates an impetus for cooperation (Dinar et al. 2010).
Wednesday September 26th at 12:30pm
Not 'Won't you be my neighbor?' but 'Should you be my neighbor?'
Animals in flocks, schools and other social networks can learn from their peers. But how many other individuals should they pay attention to? The number of individuals you attend to and the numbers they attend to affect the speed and accuracy with which information can be gathered. On the other hand, cognitive limits put constraints on the space of possible strategies. I will present early results from the relative merits of different information gathering strategies, as well as a potential metric for the informational burden of a network.
Wednesday October 3rd at 12:30pm
Dealing with death. How different mortality scenarios impact predators with complex life history in a size-structured community
Anieke van Leeuwen
The mystery posed by the lack of recovery after marine predator population collapses has evoked many hypotheses to explain this phenomenon. Fisheries impact complex ecosystems, formed by the integrated dynamics at small scales. By the analysis of a size-structured predator-prey model some of the most popular explanations for the lack of recovery of marine predator populations are tested. This modeling approach allows to link processes at the individual level to population dynamics.
We look at the impact on predator-prey dynamics from additional mortality in the predator egg stage, from cannibalism, additional mortality of adult prey and additional mortality of predators of large sizes. Contrary to intuition we find that increased mortality of predator eggs may expedite prey population extinction. Cannibalism results in larger predators and increased potential for coexistence of predator and prey. Additional mortality of large-sized prey individuals has little effect, but such mortality imposed on predators leads to the occurrence of bistability as well as predator extinction.
Wednesday October 10th at 12:30pm
Will rapid deforestation prevent endemic birds from responding to climate change in Southeast Asia?
It is unclear whether deforestation or climate change will cause more tropical bird extinctions. Here we report on the first effort to combine fine-scale climatic and dynamic land cover models to forecast vulnerability of tropical species. We sampled bird communities on four mountains across three seasons in Lore Lindu National Park, Sulawesi, Indonesia (a globally-important hotspot of avian endemism), to characterize relationships between elevation and abundance. We compared the relative impacts of climate change (projected using an ensemble of global climate models) and deforestation (based on historical rates) on abundance for two middle- and two high-elevation endemic species. Future forest area was projected under two land-use change scenarios − one assuming current deforestation rates, another assuming a 50% reduction in deforestation. Potential climate-change-induced range shifts were simulated by shifting species’ abundance distributions upslope using a locally measured adiabatic lapse rate of –6.8 °C per 1,000 m of elevation gained. Lore Lindu National Park lost 11.8% of its forest area from 2000 to 2010 and Sulawesi as a whole lost 10.8%. Global climate models forecast that Central Sulawesi may warm by 0.7–0.9 °C by 2050 (for low- and high-emissions scenarios), which could translate into a lapse-rate-linked range shift of approximately 100 m upward. Our predictions suggest that high-elevation species will be buffered from deforestation by their isolated ranges, but potentially face steep population declines from climate change (by as much as 51%). Middle-elevation species are predicted to undergo moderate declines from half-rate deforestation or climate change (11–13% reductions), while deforestation at the current rate, or climate change combined with deforestation, is predicted to cause larger declines of 16–25%. If species are to track preferred climates, they will need large areas of remnant forest, which are unlikely to remain if current deforestation patterns continue. The biological richness and rapid deforestation now occurring inside Lore Lindu National Park emphasizes the need for increased enforcement of illegal clearing in the park. Further, our results indicate that climate change is a potentially serious threat to high-elevation endemics in Central Sulawesi. These findings are likely to be applicable to many other upland tropical sites where deforestation is encroaching from below and climate change is stressing high-elevation species.
Wednesday October 17th at 2:00pm
Implications of the spatial dynamics of fire spread for the bistability of savanna and forest
Recent analyses of the global distribution of the savanna biome suggest that, depending on climate, savanna can be a stable and distinct alternative to forest, maintained by fire. Modeling work – based largely on simple models of tree populations interacting with mean-field approximations of critical behaviors in fire spread – suggests that the threshold response of fire spread to fuel-layer continuity could provide a plausible mechanism for maintaining the bistability of savanna and forest. Although empirically well supported, these modeling efforts have largely failed to explicitly take into account the spatial dynamics of fire spread. In this work (with C. Staver and S. Levin), I explore a simple spatial model providing a plausible mechanism for the bistablity of forest and savanna.
Wednesday October 24th at 12:30pm
Bigger (group size) is not always better when maximizing collective accuracy
Balancing the costs and benefits of group living often results in an optimal finite group size that in theory maximizes individual fitness. When considering only the accuracy of collective decision-making, however, it is generally agreed that larger group sizes are always better (e.g. many wrong hypothesis, the Condorcet jury theorem, Galton's vox populi). But this assumes that individuals in a group contribute statistically uncorrelated information, which will rarely be true in natural systems. Explicitly considering the correlation of information can lead to interesting results, and here I will show one result: how finite group sizes can be optimal when maximizing decision accuracy.
Wednesday October 31st at 12:30pm
Modeling global marine food-systems
James Watson, Erol Ackay
Gauging how much food we will have to eat in the future is an important task. But, the problem is complex; in the future the potential supply of food may diminish (perhaps due to climate change) and, with a growing global population, demand is likely to go up. Further, between supply and demand is the redistribution of food by markets. Here we present ongoing work to develop a modeling system for projecting food security into the future. For now our work focuses on marine-based food: i.e. fish. James will present some early results from a size-based community model that uses estimates of zooplankton biomass, produced from an Earth System Model run at GFDL, to simulate the dynamics of fish, or in terms of the food system -- potential food. This model is global, estimating the potential supply of marine-based food for all countries. Erol will then describe initial steps towards defining a global market model. We are developing a game theoretic toy-model of global markets, where we use Cournot-Nash equilibria to estimate the magnitude of fishing effort and trade between countries. With this modeling system we aim to project the winners and losers, in terms of marine-based food security, under different climate and anthropogenic scenarios… but this is a long way off. For now we aim to show some initial results, and elicit advice as to our methods.
Wednesday November 7th at 12:30pm
Modeling within-host HIV and Hepatitis C coinfection
In addition to causing its own global morbidity and mortality, HIV is changing the dynamics of other pathogens at both the population and within-host levels. One such pathogen is Hepatitis C Virus (HCV); coinfection with HIV and HCV is becoming increasingly common and problematic, particularly in communities of injecting drug users. HIV affects clearance rates and prognosis of HCV, while there is also evidence that HCV affects progression rates from HIV to full-blown AIDS. I will be presenting some preliminary work on modeling how HIV can affect the within-host dynamics of HCV spread, and specifically how HIV infection may upset sustained virologic response to treatment for HCV.
Wednesday November 14th at 12:30pm
The diffusion of public goods, reciprocity, and population structure in microbes
Microbial cells in biofilms cooperate through the secretion of public goods, such as digestive enzymes and iron chelators. In many cases, either the public good or its product can diffuse away from the producing cell. This cooperative behavior is subject to exploitation by non-producing cells.
Such a strategy confers a density-dependent benefit on the producing population, but is susceptible to exploitation by non-producing mutants and neighboring populations. Reciprocity can mitigate exploitation by reducing production in mixed environments. However, reciprocators are still vulnerable to exploitation near their frontiers. Futhermore, since resource access varies with proximity to the source, the degree of exploitation might be expected to vary with the degree of mixing between populations.
Can kin domains and/or reciprocity lead to coexistence of producers and non-producers in biofilms? To explore this question, we developed a model for an extracellular resource diffusing from a subpopulation of "cooperator" biofilm cells. THe resource has a finite diffusive range. Every cell divides stochastically with a particular rate, with access to the resource enhancing that rate and production of the resource reducing it. We modeled reciprocity through positive feedback: cooperators produce a basal level of the resource, plus a higher rate that scales with local concentration.
Surprisingly, we find that diffusion of public goods completely prevents coexistence in a producer/non-producer model. Reciprocity is not sufficient to recover coexistence. Furthermore, population dynamics are insensitive to spatial conformation. Instead, the allele frequencies evolve like those of a non-spatial population subject to weak selection and neutral drift. Our results suggest taht any fitness advantage conferred by production of diffusible public goods must occur early in the biofilm life cycle, when total population size is growing.
Wednesday November 21st at 12:30pm
No lab tea
Wednesday November 28th at 12:30pm
Leadership and the dynamics of information transfer in fish schools
One benefit of living in a group is the ability to make use of distributed sensing capabilities, which may aid in both predator avoidance and resource acquisition. In order for individuals to take advantage of these capabilities, information must be reliably transmitted across the group. Unraveling the process of information transfer is key to our understanding of collective behavior. One way to probe this process is by introducing a perturbation into a group and tracking its spread. I will present some results of an experiment with fish schools, in which a small proportion of “informed” fish, trained to respond to a stimulus, were introduced into a naïve group. This setup allows us to test theoretical predictions about how the ability of informed fish to lead the group scales with their number. Moreover, by using clear behavioral transitions as a proxy for information, we can explicitly track the spread of this “information” through the group. Simplifying the problem in this way allows for a direct comparison of models for the interaction network over which this information propagates. Our results suggest that this network is specifically visual, in contrast to most simulation models which assume either metric (all individuals within a certain distance) or topological (nearest N neighbors) interaction ranges.
Wednesday December 5th at 12:30pm
Understanding the future through unlocking the past: historical data and climate change inference
A general lack of long-term distributional data continues to be a stumbling block in our ability to detect, track, and predict range shifts in response to ongoing environmental change. Historical data derived from museums, field notes, and other archived sources provide past occurrences, yet comparisons between historic and contemporary data may yield biased inferences on how species resopnd to climate change unless detectability and mismatched methodologies are taken into account. Statistical techniques, such as occupancy models, can unlock the potential in historical occurrence data, as demonstrated by the evaluation of range shifts and community changes for the bird community of the Sierra Nevada in California over the last 100 years. This analysis of changes over a century reveals high heterogeneity in species' movements relating strongly to both temperature and precipitation changes over the same time period. From a community perspective, site-level species diversity has generally decreased over the last century, particularly due to richness declines at elevational extremes. Altogether, this analysis of historical occurrence data provides strong evidence of how climate change has already begun affecting bird ranges and communities with important implications for the future.
Monday December 10th at 11:00am
Tropical carbon sink depends on N2 fixation and biodiversity
Forests contribute a significant portion of the land carbon sink, but their future ability to sequester CO2 may be constrained by nitrogen, a major plant-limiting nutrient. Many tropical forests possess tree species capable of fixing atmospheric N2, but it is unclear whether this functional group can supply the nitrogen needed as forest carbon pools recover following disturbance and/or land-use transition. In this Lab Tea, I will identify a powerful feedback mechanism in which N2 fixation can correct ecosystem-scale deficiencies in nitrogen that emerge during periods of rapid carbon recovery in tropical forests. I will discuss results from field-work across 300 years of forest recovery from land use in Panama coupled with a theoretical model of plant-soil carbon and nitrogen dynamics. The findings indicate that the tropical CO2 sink depends on the presence and diversity of N2-fixing tree species, and offer a mechanism by which the cycles of carbon and nitrogen interact strongly with land-use transitions.
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