More pests but less pesticide applications: Ambivalent effect of landscape complexity on conservation biological control.

In agricultural landscapes, the amount and organization of crops and semi-natural habitats (SNH) have the potential to promote a bundle of ecosystem services due to their influence on ecological community at multiple spatio-temporal scales. SNH are relatively undisturbed and are often source of complementary resources and refuges, therefore supporting more diverse and abundant natural pest enemies. However, the nexus of SNH proportion and organization with pest suppression is not trivial. It is thus crucial to understand how the behavior of pest and natural enemy species, the underlying landscape structure, and their interaction, may influence conservation biological control (CBC). Here, we develop a generative stochastic landscape model to simulate realistic agricultural landscape compositions and configurations of fields and linear elements. Generated landscapes are used as spatial support over which we simulate a spatially explicit predator-prey dynamic model. We find that increased SNH presence boosts predator populations by sustaining high predator density that regulates and keeps pest density below the pesticide application threshold. However, predator presence over all the landscape helps to stabilize the pest population by keeping it under this threshold, which tends to increase pest density at the landscape scale. In addition, the joint effect of SNH presence and predator dispersal ability among hedge and field interface results in a stronger pest regulation, which also limits pest growth. Considering properties of both fields and linear elements, such as local structure and geometric features, provides deeper insights for pest regulation; for example, hedge presence at crop field boundaries clearly strengthens CBC. Our results highlight that the integration of species behaviors and traits with landscape structure at multiple scales is necessary to provide useful insights for CBC.

Discussion on "Competition on Spatial Statistics for Large Datasets".

We discuss the methods and results of the RESSTE team in the competition on spatial statistics for large datasets. In the first sub-competition, we implemented block approaches both for the estimation of the covariance parameters and for prediction using ordinary kriging. In the second sub-competition, a two-stage procedure was adopted. In the first stage, the marginal distribution is estimated neglecting spatial dependence, either according to the flexible Tuckey g and h distribution or nonparametrically. In the second stage, estimation of the covariance parameters and prediction are performed using Kriging. Vecchias's approximation implemented in the GpGp package proved to be very efficient. We then make some propositions for future competitions. SUPPLEMENTARY INFORMATION: The online version supplementary material available at 10.1007/s13253-021-00462-2.

Prediction of regional wildfire activity in the probabilistic Bayesian framework of Firelihood.

Modeling wildfire activity is crucial for informing science-based risk management and understanding the spatiotemporal dynamics of fire-prone ecosystems worldwide. Models help disentangle the relative influences of different factors, understand wildfire predictability, and provide insights into specific events. Here, we develop Firelihood, a two-component, Bayesian, hierarchically structured, probabilistic model of daily fire activity, which is modeled as the outcome of a marked point process: individual fires are the points (occurrence component), and fire sizes are the marks (size component). The space-time Poisson model for occurrence is adjusted to gridded fire counts using the integrated nested Laplace approximation (INLA) combined with the stochastic partial differential equation (SPDE) approach. The size model is based on piecewise-estimated Pareto and generalized Pareto distributions, adjusted with INLA. The Fire Weather Index (FWI) and forest area are the main explanatory variables. Temporal and spatial residuals are included to improve the consistency of the relationship between weather and fire occurrence. The posterior distribution of the Bayesian model provided 1,000 replications of fire activity that were compared with observations at various temporal and spatial scales in Mediterranean France. The number of fires larger than 1 ha across the region was coarsely reproduced at the daily scale, and was more accurately predicted on a weekly basis or longer. The regional weekly total number of larger fires (10-100 ha) was predicted as well, but the accuracy degraded with size, as the model uncertainty increased with event rareness. Local predictions of fire numbers or burned areas also required a longer aggregation period to maintain model accuracy. The estimation of fires larger than 1 ha was also consistent with observations during the extreme fire season of the 2003 unprecedented heat wave, but the model systematically underrepresented large fires and burned areas, which suggests that the FWI does not consistently rate the actual danger of large fire occurrence during heat waves. Firelihood enabled a novel analysis of the stochasticity underlying fire hazard, and offers a variety of applications, including fire hazard predictions for management and projections in the context of climate change.

What Patients Can Tell Us: Topic Analysis for Social Media on Breast Cancer.

BACKGROUND: Social media dedicated to health are increasingly used by patients and health professionals. They are rich textual resources with content generated through free exchange between patients. We are proposing a method to tackle the problem of retrieving clinically relevant information from such social media in order to analyze the quality of life of patients with breast cancer. OBJECTIVE: Our aim was to detect the different topics discussed by patients on social media and to relate them to functional and symptomatic dimensions assessed in the internationally standardized self-administered questionnaires used in cancer clinical trials (European Organization for Research and Treatment of Cancer [EORTC] Quality of Life Questionnaire Core 30 [QLQ-C30] and breast cancer module [QLQ-BR23]). METHODS: First, we applied a classic text mining technique, latent Dirichlet allocation (LDA), to detect the different topics discussed on social media dealing with breast cancer. We applied the LDA model to 2 datasets composed of messages extracted from public Facebook groups and from a public health forum (cancerdusein.org, a French breast cancer forum) with relevant preprocessing. Second, we applied a customized Jaccard coefficient to automatically compute similarity distance between the topics detected with LDA and the questions in the self-administered questionnaires used to study quality of life. RESULTS: Among the 23 topics present in the self-administered questionnaires, 22 matched with the topics discussed by patients on social media. Interestingly, these topics corresponded to 95% (22/23) of the forum and 86% (20/23) of the Facebook group topics. These figures underline that topics related to quality of life are an important concern for patients. However, 5 social media topics had no corresponding topic in the questionnaires, which do not cover all of the patients' concerns. Of these 5 topics, 2 could potentially be used in the questionnaires, and these 2 topics corresponded to a total of 3.10% (523/16,868) of topics in the cancerdusein.org corpus and 4.30% (3014/70,092) of the Facebook corpus. CONCLUSIONS: We found a good correspondence between detected topics on social media and topics covered by the self-administered questionnaires, which substantiates the sound construction of such questionnaires. We detected new emerging topics from social media that can be used to complete current self-administered questionnaires. Moreover, we confirmed that social media mining is an important source of information for complementary analysis of quality of life.

Index for Predicting Insurance Claims from Wind Storms with an Application in France.

For insurance companies, wind storms represent a main source of volatility, leading to potentially huge aggregated claim amounts. In this article, we compare different constructions of a storm index allowing us to assess the economic impact of storms on an insurance portfolio by exploiting information from historical wind speed data. Contrary to historical insurance portfolio data, meteorological variables show fewer nonstationarities between years and are easily available with long observation records; hence, they represent a valuable source of additional information for insurers if the relation between observations of claims and wind speeds can be revealed. Since standard correlation measures between raw wind speeds and insurance claims are weak, a storm index focusing on high wind speeds can afford better information. A storm index approach has been applied to yearly aggregated claim amounts in Germany with promising results. Using historical meteorological and insurance data, we assess the consistency of the proposed index constructions with respect to various parameters and weights. Moreover, we are able to place the major insurance events since 1998 on a broader horizon beyond 40 years. Our approach provides a meteorological justification for calculating the return periods of extreme-storm-related insurance events whose magnitude has rarely been reached.

Breast cancer and quality of life: medical information extraction from health forums.

Internet health forums are a rich textual resource with content generated through free exchanges among patients and, in certain cases, health professionals. We tackle the problem of retrieving clinically relevant information from such forums, with relevant topics being defined from clinical auto-questionnaires. Texts in forums are largely unstructured and noisy, calling for adapted preprocessing and query methods. We minimize the number of false negatives in queries by using a synonym tool to achieve query expansion of initial topic keywords. To avoid false positives, we propose a new measure based on a statistical comparison of frequent co-occurrences in a large reference corpus (Web) to keep only relevant expansions. Our work is motivated by a study of breast cancer patients' health-related quality of life (QoL). We consider topics defined from a breast-cancer specific QoL-questionnaire. We quantify and structure occurrences in posts of a specialized French forum and outline important future developments.