Model averaging,missing data and multiple imputation: a case study for behavioural ecology

Model averaging, specifically information theoretic approaches based on Akaike’s information criterion (IT-AIC approaches), has had a major influence on statistical practices in the field of ecology and evolution. However, a neglected issue is that in common with most other model fitting approaches, IT-AIC methods are sensitive to the presence of missing observations. The commonest way of handling missing data is the complete-case analysis (the complete deletion from the dataset of cases containing any missing values). It is well-known that this results in reduced estimation precision (or reduced statistical power), biased parameter estimates; however, the implications for model selection have not been explored. Here we employ an example from behavioural ecology to illustrate how missing data can affect the conclusions drawn from model selection or based on hypothesis testing. We show how missing observations can be recovered to give accurate estimates for IT-related indices (e.g. AIC and Akaike weight) as well as parameters (and their standard errors) by utilizing ‘multiple imputation’. We use this paper to illustrate key concepts from missing data theory and as a basis for discussing available methods for handling missing data. The example is intended to serve as a practically oriented case study for behavioural ecologists deciding on how to handle missing data in their own datasets and also as a first attempt to consider the problems of conducting model selection and averaging in the presence of missing observations.     

PCDDs, PCDFs emission control by dry scrubbing system

Presently, in Japan there are no limitations on the emission of PCDDs or PCDFs, but in order to study the feasibility of dry type air pollution control, a pilot plant was constructed in 1988 and the removal efficiencies for PCDDs, acid gas and heavy metals were measured.At the same time PCDDs concentration was compared with that of a previously installed electrostatic precipitator (ESP) plus wet scrubber line.In this paper, the following two items are reported.

1. (1) The difference in the amounts of PCDDs and PCDFs produced due to differences in gas temperature and retention time in ESP and fabric filter (FF).

2. (2) Removal efficiencies of PCDDs and PCDFs of fabric filter.

PCDDs concentration, generally 100–200 ng/Nm³ at the boiler outlet (ESP inlet and/or Quench Reactor (QR) inlet), increased several times at the ESP outlet, but it showed almost no increase at the QR outlet due to a sudden temperature drop. The temperature was 280–310°C, and the gas retention time was 12 sec. during passage through ESP so that it is thought that PCDD was formed under these conditions.On the other hand, a removal efficiency of approx. 90% was obtained with the fabric filter, and the PCDD at the bag outlet was at a sufficiently low level.     

Spatiotemporal classification of environmental monitoring data in the Yeongsan River basin, Korea, using self-organizing maps

Environmental monitoring data for planning, implementing and evaluating the Total Maximum Daily Loads (TMDL) management system have been measured at about 8-day intervals in a number of rivers in Korea since 2004. In the present study, water quality parameters such as Suspended Solids (SS), Biochemical Oxygen Demand (BOD), Dissolved Oxygen (DO), Total Nitrogen (TN), and Total Phosphorus (TP) and the corresponding runoff were collected from six stations in the Yeongsan River basin for six years and transformed into monthly mean values. With the primary objective to understand spatiotemporal characteristics of the data, a methodologically systematic application of a Self-Organizing Map (SOM) was made. The SOM application classified the environmental monitoring data into nine clusters showing exclusively distinguishable patterns. Data frequency at each station on a monthly basis identified the spatiotemporal distribution for the first time in the study area. Consequently, the SOM application provided useful information that the sub-basin containing a metropolitan city is associated with deteriorating water quality and should be monitored and managed carefully during spring and summer for water quality improvement in the river basin.     

Mechanism of Depolymerization Reaction of Polyethylene Terephthalate: Experimental and Theoretical Studies

The depolymerization reaction of polyethylene terephthalate (PET) was analyzed on the basis of both experimental data and numerical data obtained from quantum chemistry calculations. Various alcohols were used as solvents in the search for the reaction mechanism. The activation energy obtained from the optimized structures of the ground state and the transitional state was compared with our experimental data. Based on these comparisons, it was found that the formation of rings from alkoxyl groups and carbonyl carbon atoms at the transitional state is highly important when the ester bonds of PET are broken. Moreover, the most appropriate alcohol was suggested for the depolymerization reaction.     

Characterizing kinetics of transport and transformation of selenium in water-sediment microcosm free from selenium contamination using a simple mathematical model

This study developed a seven-compartment model for predicting the fate of selenium (Se) in an aquatic environment containing a water-sediment boundary. Speciation of Se in water-sediment microcosms under microaerobic conditions was measured to evaluate first-order kinetics of Se transportation and transformation. The microcosm consisted of a 10-ml solution containing 1mM soluble Se as selenate (Se6+) or selenite (Se4+) and 8 g wet sediment that was free from Se contamination, sampled from the Senri, Yamato, or Yodo Rivers in Osaka, Japan. Stepwise reaction coefficients describing transportation and transformation were determined using an inverse method on this model which includes: selenate (Se(W)6+) and selenite (Se(W)4+) in ponded water; selenate (Se(S)6+) and selenite (Se(S)4+), elemental Se (Se0), organic Se (Se2-) in sediment; and gaseous Se (DMSe). During this 1-month experiment, soluble Se was transported from ponded water to the sediment and Se was transformed sequentially to other Se species through biochemical reactions. Experimental and kinetic analyses indicated quantitatively that the Yamato River microcosm, with its high organic matter content, had a high adsorption rate of soluble Se. The Yodo River microcosm had a low adsorption rate for Se6+ and a low Se reduction rate. The Senri River microcosm had an apparent high volatilization rate of DMSe. The model developed in this study is extremely useful for predicting fate of Se in aquatic environment in the field.     

Polychlorinated dibenzo-p-dioxins and dibenzofurans in commercial diphenyl ether herbicides,and in freshwater fish collected from the application area

This paper describes the analyses of polychlorinated dibenzodioxins, dibenzofurans, and related compounds in diphenyl ether herbicides, CNP, NIP, and X-52. The levels of tetra-, penta-CDDs, and tetra-CDFs were about 150, 30, and 15 ppm, respectively in CNP formulations. NIP and X-52 contained lower chlorine alalogs. Isomer distribution of PCDDs in CNP agreed with an estimate from the reaction of chlorophenols. Freshwater fish samples heavily contaminated with CNP in the application season also contained 0.2 ppb of 1,3,6,8-tetra-CDD.