I spent Thursday afternoon once again hustling from one talk to another, with Organized Oral Session #47 (Universal Senescence? New Theories and Experimental Approaches Across the Tree of Life) being my primary focus. The writings of George C. Williams and his ingenious antagonistic pleiotropy hypothesis explaining aging have always fascinated me. Like a lot of the ideas forwarded by Williams, it seems to me that his theory explaining senescence has yet to be comprehensively tested, so I was hoping this session would provide the latest in research into aging. Annette Baudisch provided an introductory talk for the session (“When and why senescence evolves“), and profiled the forefathers of senescence theory. She suggested that Bill Hamilton was the first to theoretically explain the differential impacts of selection pressures at early and advanced ages, and that Peter Medawar pushed Hamilton’s logic further by suggesting that deleterious mutations that manifest themselves later in life are more likely to accumulate in the gene pool. She contrasted this more stochastic theory of aging with the direct tradeoffs suggested by Williams (and apparently Thomas Kirkwood, whose work I was not familiar with), who suggested that direct selection of traits beneficial in early life allowed for the genes underlying these traits to prevail even though they prove deleterious in later life. Shripad Tuljapurkar was next and provided an overview of basic life history theory that never quite allowed me to make the connection to the various hypotheses for the existence of senescence.
The next senescence talk I caught, “Life and death in a Petersham cemetery: The demography of potentially immortal organisms“, was presented by Anne Pringle. She does fascinating work on the life history of the fungal component of Xanthoparmelia lichens. Using headstones as an easy-to-document substrate for observing the natural growth of these lichens, she has been able to uncover the life history of these fungi. What she has found is that the fungi that form the basis of lichen symbioses have extremely low survival after early establishment but can become seemingly immortal once they reach a certain size. Given that her work only spans a short period of time, it still remains to be seen whether established individuals persist for longer spans of time, but it would not surprise me if they did. This brings up an interesting question: why do some organisms appear not to senescence? The logical answer is that they do not face tradeoffs between current and future reproduction, which seems to fit the life history of these lichens. If they make it through the establishment phase (in which reproductive potential is minimal), the cost of maintenance is apparently low enough to justify ‘consistent maintenance’. Pringle mused on whether her lichens are analogous to many plant species, whose modular growth and flexible sexual/asexual reproductive modes appear to sometimes obviate senescence, but she was not conclusive.
Deborah Roach spoke about “What patterns of aging emerge from a long-term, longitudinal, study of a plant population in the wild?“, directly addressing the question of whether or not plants age. Her work on Plantago lanceolata populations field-planted and monitored over an entire decade allowed her to make some tentative speculations on how plants age. One interesting result of her work was that cohorts of individuals planted at different times failed to show a consistent life history pattern. Instead, particular years produced the same result in all cohorts, suggesting that environment rather than any developmental process dictate the survivorship and reproduction of these plants. Does this mean that the plants do not age? This is a tougher question to answer, as it is hard to pull apart the effects of genes and environment in this kind of study. It is quite possible that the processes described by Medawar or Williams do affect these plants, but environmental vagaries play such a strong role in reducing survivorship that aging is not detectable in the field. To get around this problem Roach also conducted greenhouse work that showed some evidence of senescence in older plants, in particular in how they are able to modify their specific leaf area (SLA) under shaded conditions.
Anne Bronikowski’s talk, “Senescence in ectothermic vertebrates: Peaks and valleys in the landscape of reptilian aging“, presented some interesting results. Proper studies of senescence require that the individuals studied actually reach old age, and if you study reptiles this can be a challenge. Many reptiles have been shown to live for decades (or even centuries) in captivity, so long-term datasets are needed to untangle the effects of environmental mortality from aging effects. Bronikowski has inherited data from two long-term projects, one on turtles (started in 1988) and one on garter snakes (started in 1976). Using mark-recapture techniques, these studies try to estimate the survivorship curves and absolute ages (using growth rings on scales) of these two species. Remarkably, Bronikowski has found no evidence of reproductive senescence in either dataset, suggesting that the reputation of reptiles for negligible senescence may be well-earned. What I found interesting about these findings is their relation to the data presented on fungi and plants: in these diverse species where survival to reproductive maturity is the chief challenge, tradeoffs between reproduction and survival may be negligible for adults. That many organisms maintain a constant reproductive effort throughout their adult lives is fascinating.
Unfortunately, this insight was about the only one I found of value in this session, whose theoretically-oriented talks did little to shed light on the origin of senescence. None of the theoretically- or empirically-based talks provided any real insights into the ecological factors that determine whether senescence will evolve, and I left with no real sense as to whether or not recent studies had advanced our assessment of competing hypotheses explaining aging.
To be fair, I did not see all of the senescence talks. I jumped out for three specific talks throughout the afternoon, the first being Steven Railsback’s “Foraging theory for individual-based models of pest control by birds: Tests in a model of coffee borer suppression by warblers in Jamaica“. As my previous post on his workshop makes clear, I am a big fan of Railsback’s work, so I was excited to check out this talk. I was not disappointed. Railsback delivered a fascinating presentation about his modeling work on an experimental system set up by his collaborator, Matthew D. Johnson (also from Humboldt State University). Railsback explained that this Jamaican agroecosystem contains three species groups of importance: coffee plants, coffee borer beetles, and the black-throated blue warbler. The beetles are agricultural pests on the coffee, but the warbler has the potential to act as a biocontrol agent through opportunistic feeding on bursts of beetle population. While the coffee borer beetles are not the bird’s primary source of insect food, the warblers have been shown to control outbursts of the pests, leading Railsback to ask whether changing the habitat available to the warblers can impact their effectiveness as a biocontrol agent. Empirical work in the coffee fields provided two invaluable assets to the modeling efforts: baseline data on the traits of birds and beetles as well as patterns to which the model predictions can be compared.
To better understand the dynamics of warbler predation on the beetles, an individual-based model (IBM) was developed to represent different habitat types on which beetles and warblers might interact. Four different warbler foraging strategies were incorporated into different model iterations, allowing for comparison of behavioral assumptions to real-world patterns. The warbler behaviors were based on a null model of random movement, a behavior modeled around the marginal value theory of Charnov, and two “optimal cell” models that allowed birds to move to the best patches within short and long ranges. I was intrigued that Railsback incorporated an optimal foraging behavior into his model, as the assumptions underlying such optimal models (complete and global knowledge, perfect memory) are so different from those that generally underly the behaviors modeled in IBM’s. I was not surprised that Railsback had some difficulty in literally translating marginal value theory into a simulated behavior: optimal behavioral theories often posit almost mystical abilities that are in practice sometimes impossible to program into simulated organisms (and I will leave it to you to figure out how they might be genetically encoded into real organisms!). Railsback’s best estimate of what ‘Charnov behavior’ would look like in a simulated organism might not have fully captured the spirit of optimal foraging, but this to me only indicates the unrealistic nature of optimal foraging assumptions.
Railsback and Johnson had nine patterns from field observations to which to compare their various model iterations. He discussed three of these patterns to illustrate how the models faired. More birds are observed in shaded areas, and all four model iterations were able to produce this pattern. Birds also respond to local food eruptions, and the Charnov-inspired optimal departure behavior was unable to produce this pattern. They also had observed movement distances, which all but the long-range optimal cell iteration were able to replicate. Overall, the model incorporating short-range optimal cell behavior provided the best fit to the nine observed patterns. Railsback concluded that short range directed movement is an important factor in warbler behavior, that the classical optimal foraging rules fail, and that small areas of tree and shade habitat were sufficient to allow the warbler to suppress borer outbreaks. He also showed that in these models fragmentation was unimportant. This is very interesting work, work that I hope will provoke others to consider whether optimal models actually make any sense in real systems.
Jonathan Shik presented an interesting talk during a behavior session entitled “Energy subsidies from aphid mutualists fuel establishment of Argentine ant propagules“. Argentine ants are an invasive species in North America, and have successfully displaced native ants over a wide geographical range. To better understand how Argentine ants overwhelm native colonies, Shik considered the role of diet and nutrition in their invasion strategy. Like many ants, Argentine ants protect ‘herds’ of Homopteran herbivores, who produce food for the ants in the form of secreted honeydew. In order to defend Homopteran-infested plants, the ants must be able to defend not only their herd of herbivores from predators but also protect their plant ‘territory’ from other ant colonies. In order to be successful, a new Argentine ant colony must pass through a vulnerable founding period in which the queen is supported by only a few workers. Shik was interested in how dietary choices might affect this transition. One possibility is that ants focus on protein-rich insect prey during their establishment phase (because protein is needed to build worker bodies) and then shift to the carbohydrates provided by honeydew once the colony matures (when the chief requirement is for energy). To test this theory, Shik performed lab experiments with the Argentine ants, plant hosts, and various levels of insect prey and access to Homopteran mutualists. What he found overall was that the Argentine ants depend on the Homopterans at all stages of their colony development, and that access to protein in the form of insect prey provided only a minor boost to the colonies. To get more protein when worker broods are being nurtured, the ants seem to simply up their intake of food in general rather than shifting their diet. Shik’s results suggest that the mutualistic interactions of Argentine ants are critical to their success.
I also went to the talk of Paul Bourdeau, a former fellow graduate student from Stony Brook University and now a post-doc at Michigan State University. Paul’s talk, “Predicting non-consumptive predator effects on multiple prey in a complex natural system“, explored how predator-induced behavioral shifts affect prey growth and ultimately community dynamics. Anyone who studies community ecology is well-acquainted with consumptive effects: predators eat prey, depressing their numbers. But predators can also change prey behaviors, even in fairly simple organisms. Paul discussed just such a “non-consumptive predator effect” in aquatic systems: the predatory cladoceran Bythotrephes can dramatically change the behavior of its Daphnia prey. Daphnia are pretty smart for simple creatures; not only are some species of Daphnia capable of changing their morphology in response to prey, but they also can change where they live in the water column. Daphnia who detect predator cues seek out deeper, darker, colder depths where they are less vulnerable to predation but pay a large cost in the form of reduced growth. Paul is interested in how incorporating these non-consumptive effects into the community affects its overall composition and dynamics. Currently he has information on which prey species do and do not experience non-consumptive effects, and has performed lab experiments designed to parameterize the functional response of predators on various prey species. Eventually the goal is to roll all this information into a community model, which will also have to take into account seasonal variations in the water column that moderate non-consumptive effects.