Patricia Menéndez

Human impacts on fire occurrence: a case study of hundred years of forest fires in a dry alpine valley in Switzerland

Forest fire regimes are sensitive to alterations of climate, fuel load, and ignition sources. We investigated the impact of human activities and climate on fire occurrence in a dry continental valley of the Swiss Alps (Valais) by relating fire occurrence to population and road density, biomass removal by livestock grazing and wood harvest, temperature and precipitation in two distinct periods (1904–1955 and 1956–2006) using generalized additive modeling. This study provides evidence for the role played by humans and temperature in shaping fire occurrence. The existence of ignition sources promotes fire occurrence to a certain extent only; for example, high road density tends to be related to fewer fires. Changes in forest uses within the study region seem to be particularly important. Fire occurrence appears to have been negatively associated with livestock pasturing in the forest and wood harvesting, in particular during the period 1904–1955. This study illustrates consistently how fire occurrence has been influenced by land use and socioeconomic conditions. It also suggests that there is no straightforward linear relationship between human factors and fire occurrence.

Gene Regulatory Networks from Multifactorial Perturbations Using Graphical Lasso: Application to the DREAM4 Challenge

A major challenge in the field of systems biology consists of predicting gene regulatory networks based on different training data. Within the DREAM4 initiative, we took part in the multifactorial sub-challenge that aimed to predict gene regulatory networks of size 100 from training data consisting of steady-state levels obtained after applying multifactorial perturbations to the original in silico network. Due to the static character of the challenge data, we tackled the problem via a sparse Gaussian Markov Random Field, which relates network topology with the covariance inverse generated by the gene measurements. As for the computations, we used the Graphical Lasso algorithm which provided a large range of candidate network topologies. The main task was to select the optimal network topology and for that, different model selection criteria were explored. The selected networks were compared with the golden standards and the results ranked using the scoring metrics applied in the challenge, giving a better insight in our submission and the way to improve it. Our approach provides an easy statistical and computational framework to infer gene regulatory networks that is suitable for large networks, even if the number of the observations (perturbations) is greater than the number of variables (genes).

On trend estimation under monotone Gaussian subordination with long-memory: application to fossil pollen series

Fossil pollen data from stratigraphic cores are irregularly spaced in time due to non-linear age–depth relations. Moreover, their marginal distributions may vary over time. We address these features in a nonparametric regression model with errors that are monotone transformations of a latent continuous-time Gaussian process Z(T). Although Z(T) is unobserved, due to monotonicity, under suitable regularity conditions, it can be recovered facilitating further computations such as estimation of the long-memory parameter and the Hermite coefficients. The estimation of Z(T) itself involves estimation of the marginal distribution function of the regression errors. These issues are considered in proposing a plug-in algorithm for optimal bandwidth selection and construction of confidence bands for the trend function. Some high-resolution time series of pollen records from Lago di Origlio in Switzerland, which go back ca. 20,000 years are used to illustrate the methods.

Penalized regression techniques for modeling relationships between metabolites and tomato taste attributes

The search for models which link tomato taste attributes to their metabolic profiling, is a main challenge within the breeding programs that aim to enhance tomato flavor. In this paper, we compared such models calculated by the traditional statistical approach, stepwise regression, with models obtained by the new generation of regression techniques, known as penalized regression or regularization methods. In addition, for penalized regression, different scenarios and various model selection criteria were discussed to conclude that classical crossvalidation, selects models with many superfluous variables whereas model selection criteria such as Bayesian information criterion, seem to be more suitable, when the goal is to find parsimonious models, to explain tomato taste attributes based on metabolic information. An exhaustive comparison of the discussed methodology was done for six sensory traits, showing that the most important covariates were identified by the stepwise regression as well as by some of the penalized regression methods, despite the general disagreement on the size of the regression coefficients between them. In particular, for stepwise regression the coefficients are inflated due to their high variance which is not the case with penalized regression, showing that this new methodology, can be an alternative to obtain more accurate models.

Simultaneous adjustment of bias and coverage probabilities for confidence intervals

A new method is proposed for the correction of confidence intervals when the original interval does not have the correct nominal coverage probabilities in the frequentist sense. The proposed method is general and does not require any distributional assumptions. It can be applied to both frequentist and Bayesian inference where interval estimates are desired. We provide theoretical results for the consistency of the proposed estimator, and give two complex examples, on confidence interval correction for composite likelihood estimators and in approximate Bayesian computation (ABC), to demonstrate the wide applicability of the new method. Comparison is made with the double-bootstrap and other methods of improving confidence interval coverage.

A Real-Time Property Value Index Based on Web Data

House price indices are difficult to compute because houses are essentially unique, nonreplicable goods and transactions are relatively infrequent. Furthermore, unlike in organized markets, there is considerable opacity concerning transaction prices.The traditional real estate agent coexists nowadays with Web sites, from which information can be obtained cheaply and easily. This paves the way for new approaches which tap this resource to produce timely, quality and location adjusted housing price indices. We describe one such approach and show R to be a tool of choice in every step of the implementation.

Interspecific Hybridization May Provide Novel Opportunities for Coral Reef Restoration

Climate change and other anthropogenic disturbances have created an era characterized by the inability of most ecosystems to maintain their original, pristine states, the Anthropocene. Investigating new and innovative strategies that may facilitate ecosystem restoration is thus becoming increasingly important, particularly for coral reefs around the globe which are deteriorating at an alarming rate. The Great Barrier Reef (GBR) lost half its coral cover between 1985 and 2012, and experienced back-to-back heat-induced mass bleaching events and high coral mortality in 2016 and 2017. Here we investigate the efficacy of interspecific hybridization as a tool to develop coral stock with enhanced climate resilience. We crossed two Acropora species pairs from the GBR and examined several phenotypic traits over 28 weeks of exposure to ambient and elevated temperature and pCO2. While elevated temperature and pCO2 conditions negatively affected size and survival of both purebreds and hybrids, higher survival and larger recruit size were observed in some of the hybrid offspring groups under both ambient and elevated conditions. Further, interspecific hybrids had high fertilization rates, normal embryonic development, and similar Symbiodinium uptake and photochemical efficiency as purebred offspring. While the fitness of these hybrids in the field and their reproductive and backcrossing potential remain to be investigated, current findings provide proof-of-concept that interspecific hybridization may produce genotypes with enhanced climate resilience, and has the potential to increase the success of coral reef restoration initiatives.

Amphetamine availability predicts amphetamine-related mental health admissions: A time series analysis:

Objective: Amphetamine use and availability have increased in Australia and there are concerns that this has led to more frequent hospital admissions with amphetamine-related psychosis. This study examines whether amphetamine-related admissions to mental health units are more common at times of greater amphetamine availability. Methods: We conducted an ecological study using aggregate crime and health service data for NSW, Australia, from January 2000 to March 2015. Amphetamine-related criminal incidents (arrests or cautions for possession or use) were used as an indirect measure of amphetamine availability. Semiparametric time series analysis was used to compare monthly arrest rates to monthly hospitalisation rates for (1) amphetamine abuse or dependence, (2) amphetamine-related psychosis and (3) any psychosis. Results: Amphetamine-related admissions to NSW mental health units have increased four- to fivefold since 2009 and comprised approximately 10% of all admissions to these units in early 2015. There was a significant association between arrests and amphetamine-related admissions. After adjustment for seasonal variation, this effect demonstrated a time lag of 1–2 months. There was no relationship between amphetamine arrests and overall admissions for psychosis. Conclusion: Greater amphetamine availability significantly predicts admissions for amphetamine use disorders and amphetamine-related psychosis. Better treatment strategies are needed to break the nexus between drug availability and drug-related harm.