Using between-model comparisons to fine-tune linear models of species ranges

Aim To offer an objective approach to some of the problems associated with the development of logistic regression models: how to compare different models, determination of sample size adequacy, the influence of the ratio of positive to negative cells on model accuracy, and the appropriate scale at which the hypothesis of a non‐random distribution should be tested. Location Test data were taken from Southern Africa. Methods The approach relies mainly on the use of the AUC (Area under the Curve) statistic, based on ROC (threshold Receiver Operating Characteristic) plots, for between‐model comparisons. Data for the distribution of the bont tick Amblyomma hebraeum Koch (Acari: Ixodidae) are used to illustrate the methods. Results Methods for the estimation of minimum sample sizes and more accurate hypothesis‐testing are outlined. Logistic regression is robust to the assumption that uncollected cells can be scored as negative, provided that the sample size of cells scored as positive is adequate. The variation in temperature and rainfall at localities where A. hebraeum has been collected is significantly lower than expected from a random sample of points across the data set, suggesting that within‐site variation may be an important determinant of its distribution. Main conclusions Between‐model comparisons relying on AUCs can be used to enhance objectivity in the development and refinement of logistic regression models. Both between‐site and within‐site variability should be considered as potentially important factors determining species distributions.     

Agonist binding fraction of dopamine D2/3 receptors in rat brain: A quantitative autoradiographic study

There has arisen considerable interest in the study of dopamine D_2/3 agonist binding sites by positron emission tomography (PET), based on the claim that agonist sites represent a functional subset of the total number of sites labeled by more conventional antagonist ligands. To test the basis of this claim, we used quantitative autoradiography to measure the abundance of binding sites of a dopamine D_2/3 agonist ([³H]NPA) and an antagonist ([³H]raclopride) in cryosections of rat brain. Saturation binding studies revealed that the B_max for [³H]NPA was nearly identical to that of [³H]raclopride in dorsal brain regions, but was 25% less in the ventral striatum and 56% less in the olfactory tubercle. We also tested the displacement of the two ligands by the hallucinogen LSD, which is known to have dopamine agonist properties. Whereas displacement of [³H]raclopride by increasing LSD concentrations was monophasic, displacement of [³H]NPA was biphasic, suggesting an action of LSD via a subset of dopamine D_2/3 agonist binding sites. Addition of the stable GTP analogue Gpp(NH)p to the medium abolished 90% of the [³H]NPA binding, and increased [³H]raclopride binding by 10%, with a shift to the right in the LSD competition curve, suggesting retention of endogenous dopamine in washed cryostat sections. Thus [³H]NPA and [³H]raclopride binding sites have nearly identical abundances in rat dorsal striatum, but are distinct in the ventral striatum, and with respect to their displacement by LSD.     

Predation on reef-building corals: multiscale variation in the density of three corallivorous gastropods, Drupella spp.

 Feeding aggregations of three corallivorous gastropods, Drupella cornus, D. fragum and D. rugosa, have caused considerable coral damage on reefs across the Indo-West Pacific. Distribution and abundance of these three species were explored at Lizard Island, Great Barrier Reef, to determine within-reef variations in density, and spatial relationships between Drupella and their prey corals. The scales of greatest variation were between reef habitats (combinations of exposure and depth) and individual coral colonies. Density varied 12-fold among four habitats: exposed crests (2.55/m²), exposed slopes (0.22/m²), sheltered crests (0.34/m²) and sheltered slopes (2.07/m²). Species composition also varied markedly between habitats. Individuals were highly aggregated, usually forming small clusters (<10 individuals) on live coral colonies and other substrata, and occasional large aggregations of=200 to>2000. Five basic tenets for sampling Drupella are established, based on patterns of variation in density and species composition, and small-scale habitat use. Accepted: 8 February 1999     

Neuronal activity and its links with the perception of multi-stable figures

In order to isolate the neuronal activity that relates to the making of perceptual decisions, we have made use of a perceptually ambiguous motion stimulus. This stimulus lies on the boundary between two perceptual categories that correspond to clockwise and counter-clockwise rotation of a three-dimensional figure. It consists of a two-dimensional pattern of moving dots that are capable of generating these two, distinct, three-dimensional percepts. We have studied the responses of neurons in cortical area V5/MT whilst macaque monkeys report judgements about the perceptual configuration of this stimulus. We extract a quantitative statistic called 'choice probability' that expresses the covariation of neuronal activity and perceptual choice. An analysis of choice probabilities shows that the pool of neurons involved in the perceptual decisions is a tightly constrained subset of the population of sensory neurons relevant to the perceptual task.     

Multiple states in river and lake ecosystems

Nonlinear models of ecosystem dynamics that incorporate positive feedbacks and multiple, internally reinforced states have considerable explanatory power. However, linear models may be adequate, particularly if ecosystem behaviour is primarily controlled by external processes. In lake ecosystems, internal (mainly biotic) processes are thought to have major impacts on system behaviour, whereas in rivers, external (mainly physical) factors have traditionally been emphasized. We consider the hypothesis that models that exhibit multiple states are useful for understanding the behaviour of lake ecosystems, but not as useful for understanding stream ecosystems. Some of the best-known examples of multiple states come from lake ecosystems. We review some of these examples, and we also describe examples of multiple states in rivers. We conclude that the hypothesis is an oversimplification; the importance of physical forcing in rivers does not eliminate the possibility of internal feedbacks that create multiple states, although in rivers these feedbacks are likely to include physical as well as biotic processes. Nonlinear behaviour in aquatic ecosystems may be more common than current theory indicates.     

An Investigation into the Stabilization of Diltiazem HCl Release from Matrices Made from Aged Polyox Powders

Matrices containing PEO fail to provide stable drug release profiles when stored at elevated temperatures for a period of time. The present study aims to stabilize diltiazem HCl release from matrices made from various molecular weights of polyox powders. To this end, various molecular weights of polyox with and without vitamin E (0.25, 0.5 and 1% w/w) were stored at 40°C for 0, 2, 4 and 8 weeks. The aged polyox powders were then mixed with the model drug at a ratio of 1:1 and compressed into tablets. At different time intervals, the aged polyox with vitamin E were taken out of oven and mixed with the drug (1:1 ratio) and compressed into tablets. Dissolution studies showed a significant increase in diltiazem HCl release rate to occur with increased storage time at 40°C ± 1 from tablets made from the aged polyox (no vitamin E). This was as a result of depolymerization of the aged polyox powders as compared to the fresh polyox samples. This was confirmed by differential scanning calorimetry (DSC) which showed a reduction in the melting point of the aged samples. Concentrations of vitamin E as low as 0.25% w/w was able to overcome the quick release of drug from the matrices made from aged polyox powders. DSC traces showed that the melting point of aged polyox samples containing vitamin E remained the same as that of the fresh samples. The presence of vitamin E is essential to stabilize the drug release from polyox matrices containing diltiazem HCl.Key words: depolymerization, drug release kinetics, molecular weight, polyox matrices, thermal behaviour