Model selection and model averaging in behavioural ecology: the utility of the IT-AIC framework

Behavioural ecologists often study complex systems in which multiple hypotheses could be proposed to explain observed phenomena. For some systems, simple controlled experiments can be employed to reveal part of the complexity; often, however, observational studies that incorporate a multitude of causal factors may be the only (or preferred) avenue of study. We assess the value of recently advocated approaches to inference in both contexts. Specifically, we examine the use of information theoretic (IT) model selection using Akaike’s information criterion (AIC). We find that, for simple analyses, the advantages of switching to an IT-AIC approach are likely to be slight, especially given recent emphasis on biological rather than statistical significance. By contrast, the model selection approach embodied by IT approaches offers significant advantages when applied to problems of more complex causality. Model averaging is an intuitively appealing extension to model selection. However, we were unable to demonstrate consistent improvements in prediction accuracy when using model averaging with IT-AIC; our equivocal results suggest that more research is needed on its utility. We illustrate our arguments with worked examples from behavioural experiments.     

Climate change and forests of the future: managing in the face of uncertainty.

We offer a conceptual framework for managing forested ecosystems under an assumption that future environments will be different from present but that we cannot be certain about the specifics of change. We encourage flexible approaches that promote reversible and incremental steps, and that favor ongoing learning and capacity to modify direction as situations change. We suggest that no single solution fits all future challenges, especially in the context of changing climates, and that the best strategy is to mix different approaches for different situations. Resources managers will be challenged to integrate adaptation strategies (actions that help ecosystems accommodate changes adaptively) and mitigation strategies (actions that enable ecosystems to reduce anthropogenic influences on global climate) into overall plans. Adaptive strategies include resistance options (forestall impacts and protect highly valued resources), resilience options (improve the capacity of ecosystems to return to desired conditions after disturbance), and response options (facilitate transition of ecosystems from current to new conditions). Mitigation strategies include options to sequester carbon and reduce overall greenhouse gas emissions. Priority-setting approaches (e.g., triage), appropriate for rapidly changing conditions and for situations where needs are greater than available capacity to respond, will become increasingly important in the future.     

Prenatal diagnosis of pseudo arylsulphatase a deficiency

Prenatal diagnosis was performed on a pregnancy at risk for metachromatic leukodystrophy (MLD) in a family with the pseudo arylsulphatase A deficiency trait. Extracts of cultured amniotic fluid cells were deficient in arylsulphatase A indicating that the fetus was either affected with MLD or had the benign pseudodeficiency trait. In the cerebroside sulphate loading test, the at risk cells hydrolysed sulphatide like control cultured amniotic fluid cells implying that the fetus had pseudodeficiency. The pregnancy was carried to term and a male child was delivered. Placenta, urine and fibroblasts had very low activities of arysulphatase A. However, no sulphatide could be detected in urine and growing fibroblasts responded normally in the cerebroside sulphate loading test, suggesting pseudodeficiency. At 29 months, the infant is healthy and shows no stigmata of MLD. The prediction based on the results of the cerebroside sulphate loading test on cultured amniotic fluid cells appeared to be borne out.     

Optimal foraging: Some simple stochastic models

Summary Some simple stochastic models of optimal foraging are considered. Firstly, mathematical renewal theory is used to make a general model of the combined processes of search, encounter, capture and handling. In the case where patches or prey items are encountered according to a Poisson process the limiting probability distribution of energy gain is found. This distribution is found to be normal and its mean and variance are specified. This result supports the use of Holling's disc equation to specify the rate of energy intake in foraging models. Secondly, a model based on minimization of the probability of death due to an energetic shortfall is presented. The model gives a graphical solution to the problem of optimal choices when mean and variance are related. Thirdly, a worked example using these results is presented. This example suggests that there may be natural relationships between mean and variance which make solutions to the problems of energy maximization and minimization of the probability of starvation similar. Finally, current trends in stochastic modeling of foraging behavior are critically discussed.     

A simple spatially explicit ideal-free distribution: a model and an experiment

This investigation presents a simple spatially explicit analysis of the ideal-free distribution. The traditional ideal-free distribution assumes discrete sites with definite boundaries, and predicts how many individuals should occupy each site. In contrast, the present analysis assumes that a forager’s gains gradually decline with distance from a site, and asks where in space individuals ought to be. Although many interesting situations may arise, the analysis asks how individuals should position themselves as the distance between two identical sources increases. Nash equilibrium positions should follow a pitchfork pattern as the distance between sites is increased; that is, an individual should maintain a position between two sources when they are close together but should move nearer one of the sources when they are far apart. In addition, the text describes an experimental study that parallels the theoretical analysis. The experiment supports the predicted pitchfork pattern, and provides somewhat weaker support for the predicted differences in ”individual” and ”paired” pitchforks. Received: 14 June 2000 / Revised: 20 September 2000 / Accepted: 7 October 2000     

Sorption of isoxaflutole diketonitrile degradate (DKN) and dicamba in unsaturated soil

When analyzing the sorption characteristics of weakly sorbing or labile pesticides, batch methods tend to yield a high margin of error attributable to errors in concentration measurement and to degradation, respectively. This study employs a recently developed unsaturated transient flow method to determine the sorption of isoxaflutole's herbicidally active diketonitrile degradate (DKN) and dicamba. A 20-cm acrylic column was packed with soils with varied texture that had been uniformly treated with 14C-labeled chemical.The antecedent solution herbicide in equilibrium with sorbed phase herbicide was displaced by herbicide-free water, which was infiltrated into the column. Sorption coefficients, Kd, were obtained from a plot of total herbicide concentration in the soil versus water content in the region where the antecedent solution accumulated. DKN Kd values were approximately 2-3 times (average Kd = 0.71 L kg-1) greater using the unsaturated transient flow method as compared to the batch equilibration method in clay loam (Kd = 0.33 L kg-1), but similar for the two methods in sand (0.12 vs 0.09 L kg-1) soils. Dicamba Kd values were 3 times greater using the unsaturated transient flow method as compared to the batch equilibration method in the clay loam soil (0.38 vs 0.13 L kg-1), however, the Kd values were the same for the two methods in the sand (approximately 0.06 L kg-1). This demonstrates that to determine sorption coefficients for labile hydrophilic pesticides, an unsaturated transient flow method may be a suitable alternative to the batch method. In fact, it may be better in cases where transport models have overpredicted herbicide leaching when batch sorption coefficients have been used.     

Bacterial degradation of acetone in an outdoor model stream

Diurnal variations of the acetone concentration in an outdoor model stream were measured with and without a nitrate supplement to determine if the nitrate supplement would stimulate bacterial degradation of the acetone. Acetone loss coefficients were computed from the diurnal data using a fitting procedure based on a Lagrangian particle model. The coefficients indicated that bacterial degradation of the acetone was occurring in the downstream part of the stream during the nitrate addition. However, the acetone concentrations stabilized at values considerably above the limit of detection for acetone determination, in contrast to laboratory respirometer studies where the acetone concentration decreased rapidly to less than the detection limit, once bacterial acclimation to the acetone had occurred. One possible explanation for the difference in behavior was the limited 6-hour residence time of the acetone in the model stream.     

Food resource utilization among five species of embiotocids at King Harbor,California, with preliminary estimates of caloric intake

The diets of 5 species of embiotocids (Rhacochilus vacca, Embiotoca jacksoni, Hypsurus caryi ryi, Phanerodon furcatus and Micrometrus minimus) from King Harbor, Redondo Beach, California, USA, are examined by means of a dietary survey, field feeding observations and bomb calorimetry of prey items. Each species is shown to have statistically different diets from all others, with the exception of H. caryi which apparently is a dietary intermediate between E. jacksoni and P. furcatus. Dietary overlap between E. jacksoni and H. caryi appears to have resulted in some competitive exclusion along other niche parameters. Estimated caloric intakes suggest that R. vacca has the highest caloric yield per bite and the lowest feeding rate of the species studied. Possible energetic advantages are found for diet specialization in R. vacca.