Merits of Modelling Multivariate Survival Data Using Random Effects Proportional Odds Model

Recently, there has been a great deal of interest in the analysis of multivariate survival data. In most epidemiological studies, survival times of the same cluster are related because of some unobserved risk factors such as the environmental or genetic factors. Therefore, modelling of dependence between events of correlated individuals is required to ensure a correct inference on the effects of treatments or covariates on the survival times. In the past decades, extension of proportional hazards model has been widely considered for modelling multivariate survival data by incorporating a random effect which acts multiplicatively on the hazard function. In this article, we consider the proportional odds model, which is an alternative to the proportional hazards model at which the hazard ratio between individuals converges to unity eventually. This is a reasonable property particularly when the treatment effect fades out gradually and the homogeneity of the population increases over time. The objective of this paper is to assess the influence of the random effect on the within‐subject correlation and the population heterogeneity. We are particularly interested in the properties of the proportional odds model with univariate random effect and correlated random effect. The correlations between survival times are derived explicitly for both choices of mixing distributions and are shown to be independent of the covariates. The time path of the odds function among the survivors are also examined to study the effect of the choice of mixing distribution. Modelling multivariate survival data using a univariate mixing distribution may be inadequate as the random effect not only characterises the dependence of the survival times, but also the conditional heterogeneity among the survivors. A robust estimate for the correlation of the logarithm of the survival times within a cluster is obtained disregarding the choice of the mixing distributions. The sensitivity of the estimate of the regression parameter under a misspecification of the mixing distribution is studied through simulation. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Bayesian Predictive Probability Functions for Count Data that are Subject to Misclassification

We develop three Bayesian predictive probability functions based on data in the form of a double sample. One Bayesian predictive probability function is for predicting the true unobservable count of interest in a future sample for a Poisson model with data subject to misclassification and two Bayesian predictive probability functions for predicting the number of misclassified counts in a current observable fallible count for an event of interest. We formulate a Gibbs sampler to calculate prediction intervals for these three unobservable random variables and apply our new predictive models to calculate prediction intervals for a real‐data example. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Within-tree and seasonal distribution of the pine woolly aphid Pineus boerneri on Pinus kesiya trees

1 The spatial and temporal distribution of pine woolly aphids, Pineus boerneri (Homoptera: Adelgidae) a pest of exotic pine trees in Malawi, was investigated in 5‐year‐old Pinus kesiya trees. 2 Pineus boerneri was generally found to settle and reproduce on the outer shoot‐end sections of young P. kesiya trees. There was no evidence of preference for any particular levels of the tree canopy. 3 The study suggests that the biological performance and, by implication, the pest status of P. boerneri in young P. kesiya trees can be influenced by external factors, particularly by the within‐tree and seasonal fluctuations in the level of nitrogen in the host trees, total rainfall and canopy structure. 4 These findings are discussed in relation to possible management and control of P. boerneri in Malawi.     

The northern limit of Pinus banksiana Lamb. in Canada: explaining the difference between the eastern and western distributions

Aim Present northern distribution limit of jack pine (Pinus banksiana Lamb.) follows the northern limit of continuous open boreal forest in western Canada, but not in eastern Canada where it is located further south. We tested the hypothesis that fire plays a more important role than climate in explaining the present position of the northern distribution limit of jack pine. Location An experimental jack pine plantation was set up in 1992, c. 300 km north of the present distribution limit of the species, in the Boniface river area of northern Québec (57°43′ N, 76°05′ W). Methods Climate and fire data were used to compare sites at and north of the present distribution limit of jack pine. In 2001, surviving individuals from the plantation were measured (total height, annual shoot elongation, basal diameter, and presence/absence of cones). Results Climate data from the ten weather stations used in this study did not show major differences. The northern limit of jack pine distribution is closely associated with the occurrence of fires larger than 200 ha. Survival of the planted jack pines was 31%. About 25% of the surviving pines qualified as normal, single‐stem individuals; the others were slightly uprooted and/or showed marks of erosion or foraging. Cones were produced, although no viable seeds were found. Main conclusions The low number of degree‐days above 5 °C at the plantation site could explain why the seeds were not viable. However, such climate conditions are not sufficient to prevent growth, as was shown by annual shoot elongation measurements. Most of the surviving jack pines from the Boniface river plantation are relatively healthy and follow a normal developmental programme. Low fire frequency and small fire size are amongst the main factors that prevented P. banksiana from migrating further north or east following deglaciation in northern Québec and Labrador.     

Predator–prey interactions between Dugesia gonocephala and free-living nematodes

1. Three groups of laboratory experiments clarified the role of nematodes as a potential food resource for the triclad Dugesia gonocephala. The first group measured the functional response of adult D. gonocephala feeding on juvenile or adult Caenorhabditis elegans. The feeding rates of D. gonocephala on adult and juvenile C. elegans followed a type II functional response. The maximum number of adult nematodes and juvenile nematodes eaten by a single D. gonocephala individual within 3 h was 94 and 197 nematodes, respectively. 2. A second group of microcosm experiments investigated the effect of D. gonocephala on the density and the vertical distribution of a nematode community in fine sand. The following treatments were performed: (i) microcosms with 400 nematodes and (ii) microcosms with 400 nematodes and one D. gonocephala. After 5 days, nematodes as a group, as well as the dominant species Tobrilus pellucidus and Trischistoma monohystera, showed no significant difference in vertical patterns between the treatments with and without D. gonocephala. 3. The third group of experiments determined whether grain size of the sediment (sand, fine gravel and coarse gravel) altered the ability of D. gonocephala to consume adult C. elegans. Sand and fine gravel reduced the predation effectiveness of D. gonocephala by 100%, whereas the predator consumed nematodes in coarse gravel (19 nematodes within 3 h).     

SIZE-AND AGE-CLASS SEGREGATION OF BOWHEAD WHALES SUMMERING IN NORTHERN FOXE BASIN: A PHOTOGRAMMETRIC ANALYSIS

To determine whether Hudson Bay-Foxe Basin bowhead whales segregate on the basis of age, whales summering in northern Foxe Basin, were aerially photographed in August of 1996, 1997, and 1998. Image lengths on either the negatives or contact prints were measured and total body lengths were estimated. In all three years the majority of whales photographed were ≤13.5 m long. Calves and juveniles made up 89.3%, 96.6%, and 79.3% of the total number of measured whales in 1996 (n = 28), 1997 (n = 30) and 1998 (n = 29) respectively. The number of bowheads >13.5 m, the approximate size at which females reach sexual maturity, that were photographed was directly proportional to the number of calves photographed. Our results indicate that northern Foxe Basin bowheads are part of a more widely distributed stock. Adult males and resting adult females apparently summer in another part of the range, probably northwestern Hudson Bay. Northern Foxe Basin appears to be used as a summer feeding area by cows with young-of-the-year calves and by juveniles.     

Occurrence and seasonal dynamics of Pseudodactylogyrus anguillae (Yin & Sproston) (Monogenea) in eel, Anguilla anguilla (L.), in England

Established populations of Pseudodacrylogyrus anguillae are reported from eels in England for the first time. Prevalence and abundance peak in late summer in all three localities and the parasite overwinters at low levels on eels.     

Genetic relationships among American species of the genus Prosopis (Mimosoideae, Leguminosae) inferred from ITS sequences: evidence for long-distance dispersal

Aim  The genus Prosopis includes 44 species and has a pseudoamphitropical, disjunct distribution. We aimed to determine whether American Prosopis sections arose in North or South America, and to explain the current distribution of their species on the basis of their genetic relationships.
Location  South-western USA, Mexico, Caribbean Antilles, Peru–Ecuador, central and northern Argentina, south-western Argentina (Patagonia) and Cuyo, south-western Asia and northern Africa.
Methods  Internal transcribed spacer fragments from 21 species of Prosopis were sequenced and the data were used to analyse the phylogenetic relationships using Microlobius and Mimosa as outgroups. Genetic distances were calculated to estimate the degree of divergence. Dispersal–vicariance (DIVA) analysis was conducted to help understand the biogeographical history of the genus.
Main conclusions  The sections Strombocarpa and Algarobia are not monophyletic. Prosopis argentina (section Monilicarpa ) and the species of Algarobia are included in single clade. The phylogeny, DIVA analysis, and the pattern of genetic distances indicate that the ancestral area for the American species was wide, from south-western USA to Central and northern Argentina. Successive vicariance events split this area, and long-distance dispersal episodes (perhaps mediated by birds) led to recolonizations from North to South America, and vice versa .