Lotka-Volterra competition models for sessile organisms

Markov models are widely used to describe the dynamics of communities of sessile organisms, because they are easily fitted to field data and provide a rich set of analytical tools. In typical ecological applications, at any point in time, each point in space is in one of a finite set of states (e.g., species, empty space). The models aim to describe the probabilities of transitions between states. In most Markov models for communities, these transition probabilities are assumed to be independent of state abundances. This assumption is often suspected to be false and is rarely justified explicitly. Here, we start with simple assumptions about the interactions among sessile organisms and derive a model in which transition probabilities depend on the abundance of destination states. This model is formulated in continuous time and is equivalent to a Lotka-Volterra competition model. We fit this model and a variety of alternatives in which transition probabilities do not depend on state abundances to a long-term coral reef data set. The Lotka-Volterra model describes the data much better than all models we consider other than a saturated model (a model with a separate parameter for each transition at each time interval, which by definition fits the data perfectly). Our approach provides a basis for further development of stochastic models of sessile communities, and many of the methods we use are relevant to other types of community. We discuss possible extensions to spatially explicit models.     

Quantifying the complexities of Saccharomyces cerevisiae's ecosystem engineering via fermentation

The theory of niche construction suggests that organisms may engineer environments via their activities. Despite the potential of this phenomenon being realized by Darwin, the capability of niche construction to generally unite ecological and evolutionary biology has never been empirically quantified. Here I quantify the fitness effects of Saccharomyces cerevisiae's ecosystem engineering in a natural ferment in order to understand the interaction between ecological and evolutionary processes. I show that S. cerevisiae eventually dominates in fruit niches, where it is naturally initially rare, by modifying the environment through fermentation (the Crabtree effect) in ways which extend beyond just considering ethanol production. These data show that an additional cause of S. cerevisiae's competitive advantage over the other yeasts in the community is due to the production of heat via fermentation. Even though fermentation is less energetically efficient than respiration, it seems that this trait has been selected for because its net effect provides roughly a 7% fitness advantage over the other members of the community. These data provide an elegant example of niche construction because this trait clearly modifies the environment and therefore the selection pressures to which S. cerevisiae, and other organisms that access the fruit resource, including humans, are exposed to.     

Planning for persistence in marine reserves: a question of catastrophic importance.

Large-scale catastrophic events, although rare, lie generally beyond the control of local management and can prevent marine reserves from achieving biodiversity outcomes. We formulate a new conservation planning problem that aims to minimize the probability of missing conservation targets as a result of catastrophic events. To illustrate this approach we formulate and solve the problem of minimizing the impact of large-scale coral bleaching events on a reserve system for the Great Barrier Reef, Australia. We show that by considering the threat of catastrophic events as part of the reserve design problem it is possible to substantially improve the likely persistence of conservation features within reserve networks for a negligible increase in cost. In the case of the Great Barrier Reef, a 2% increase in overall reserve cost was enough to improve the long-run performance of our reserve network by >60%. Our results also demonstrate that simply aiming to protect the reefs at lowest risk of catastrophic bleaching does not necessarily lead to the best conservation outcomes, and enormous gains in overall persistence can be made by removing the requirement to represent all bioregions in the reserve network. We provide an explicit and well-defined method that allows the probability of catastrophic disturbances to be included in the site selection problem without creating additional conservation targets or imposing arbitrary presence/absence thresholds on existing data. This research has implications for reserve design in a changing climate.     

Role of plant enemies in the forestry of indigenous vs. nonindigenous pines

Plantations of rapidly growing trees are becoming increasingly common because the high productivity can enhance local economies, support improvements in educational systems, and generally improve the quality of life in rural communities. Landowners frequently choose to plant nonindigenous species; one rationalization has been that silvicultural productivity is enhanced when trees are separated from their native herbivores and pathogens. The expectation of enemy reduction in nonindigenous species has theoretical and empirical support from studies of the enemy release hypothesis (ERH) in the context of invasion ecology, but its relevance to forestry has not been evaluated. We evaluated ERH in the productive forests of Galicia, Spain, where there has been a profusion of pine plantations, some with the indigenous Pinus pinaster, but increasingly with the nonindigenous P. radiata. Here, one of the most important pests of pines is the indigenous bark beetle, Tomicus piniperda. In support of ERH, attacks by T. piniperda were more than twice as great in stands of P. pinaster compared to P. radiata. This differential held across a range of tree ages and beetle abundance. However, this extension of ERH to forestry failed in the broader sense because beetle attacks, although fewer on P. radiata, reduced productivity of P. radiata more than that of P. pinaster (probably because more photosynthetic tissue is lost per beetle attack in P. radiata). Productivity of the nonindigenous pine was further reduced by the pathogen, Sphaeropsis sapinea, which infected up to 28% of P. radiata but was absent in P. pinaster. This was consistent with the forestry axiom (antithetical to ERH) that trees planted "off-site" are more susceptible to pathogens. Fungal infections were positively correlated with beetle attacks; apparently T. piniperda facilitates S. sapinea infections by creating wounds and by carrying fungal propagules. A globally important component in the diminution of indigenous flora has been the deliberate large-scale propagation of nonnative trees for silviculture. At least for Pinus forestry in Spain, reduced losses to pests did not rationalize the planting of nonindigenous trees. There would be value in further exploration of relations between invasion ecology and the forestry of nonindigenous trees.     

Allowable levels of take for the trade in Nearctic songbirds

The take of Nearctic songbirds for the caged-bird trade is an important cultural and economic activity in Mexico, but its sustainability has been questioned. We relied on the theta-logistic population model to explore options for setting allowable levels of take for 11 species of passerines that were subject to legal take in Mexico in 2010. Because estimates of population size necessary for making-periodic adjustments to levels of take are not routinely available, we examined the conditions under which a constant level of take might contribute to population depletion (i.e., a population below its level of maximum net productivity). The chance of depleting a population is highest when levels of take are based on population sizes that happen to be much lower or higher than the level of maximum net productivity, when environmental variation is relatively high and serially correlated, and when the interval between estimation of population size is relatively long (> or = 5 years). To estimate demographic rates of songbirds involved in the Mexican trade we relied on published information and allometric relationships to develop probability distributions for key rates, and then sampled from those distributions to characterize the uncertainty in potential levels of take. Estimates of the intrinsic rate of growth (r) were highly variable, but median estimates were consistent with those expected for relatively short-lived, highly fecund species. Allowing for the possibility of nonlinear density dependence generally resulted in allowable levels of take that were lower than would have been the case under an assumption of linearity. Levels of take authorized by the Mexican government in 2010 for the 11 species we examined were small in comparison to relatively conservative allowable levels of take (i.e., those intended to achieve 50% of maximum sustainable yield). However, the actual levels of take in Mexico are unknown and almost certainly exceed the authorized take. Also, the take of Nearctic songbirds in other Latin American and Caribbean countries ultimately must be considered in assessing population-level impacts.