Understanding habitat-level variation in community structure provides an informed basis for natural resources’ management. Reef fishes are a major component of tropical marine biodiversity, but their abundance and distribution are poorly assessed beyond conventional SCUBA diving depths. Based on a baited-video survey of fish assemblages in Southwestern Atlantic’s most biodiverse region we show that species composition responded mainly to the two major hard-bottom megahabitats (reefs and rhodolith beds) and to the amount of light reaching the bottom. Both megahabitats encompassed typical reef fish assemblages but, unexpectedly, richness in rhodolith beds and reefs was equivalent. The dissimilar fish biomass and trophic structure in reefs and rhodolith beds indicates that these systems function based on contrasting energy pathways, such as the much lower herbivory recorded in the latter. Rhodolith beds, the dominant benthic megahabitat in the tropical Southwestern Atlantic shelf, play an underrated role as fish habitats, and it is critical that they are considered in conservation planning.

Tropical reefs are declining rapidly due to climate changes and local stressors such as water quality deterioration and overfishing. The so-called marginal reefs sustain significant coral cover and growth but are dominated by fewer species adapted to suboptimal conditions to most coral species. However, the dynamics of marginal systems may diverge from that of the archetypical oligotrophic tropical reefs, and it is unclear whether they are more or less susceptible to anthropogenic stress. Here, we present the largest (100 fixed quadrats at five reefs) and longest time series (13 years) of benthic cover data for Southwestern Atlantic turbid zone reefs, covering sites under contrasting anthropogenic and oceanographic forcing. Specifically, we addressed how benthic cover changed among habitats and sites, and possible dominance-shift trends. We found less temporal variation in offshore pinnacles’ tops than on nearshore ones and, conversely, higher temporal fluctuation on offshore pinnacles’ walls than on nearshore ones. In general, the Abrolhos reefs sustained a stable coral cover and we did not record regional-level dominance shifts favoring other organisms. However, coral decline was evidenced in one reef near a dredging disposal site. Relative abundances of longer-lived reef builders showed a high level of synchrony, which indicates that their dynamics fluctuate under similar drivers. Therefore, changes on those drivers could threaten the stability of these reefs. With the intensification of thermal anomalies and land-based stressors, it is unclear whether the Abrolhos reefs will keep providing key ecosystem services. It is paramount to restrain local stressors that contributed to coral reef deterioration in the last decades, once reversal and restoration tend to become increasingly difficult as coral reefs degrade further and climate changes escalate.

Remote-sensing processes based on unmanned aerial vehicles (UAV) have opened up new possibilities to both map and extract individual plant parameters. This is mainly due to the high spatial data resolution and acquisition flexibility of UAVs. Among the possible plant-related metrics is the leaf area index (LAI), which has already been successfully estimated in agronomy and forestry studies using the traditional normalized difference vegetation index from multispectral data or using hyperspectral data. However, the LAI has not been estimated in chestnut trees, and few studies have explored the use of multiple vegetation indices to improve LAI estimation from aerial imagery acquired by UAVs. This study uses multispectral UAV-based data from a chestnut grove to estimate the LAI for each tree by combining vegetation indices computed from different segments of the electromagnetic spectrum with geometrical parameters. Machine-learning techniques were evaluated to predict LAI with robust algorithms that consider dimensionality reduction, avoiding over-fitting, and reduce bias and excess variability. The best achieved coefficient of determination ($R^2$) value of 85%, which shows that the biophysical and geometrical parameters can explain the LAI variability. This result proves that LAI estimation is improved when using multiple variables instead of a single vegetation index. Furthermore, another significant contribution is a simple, reliable, and precise model that relies on only two variables to estimate the LAI in individual chestnut trees.

The reported number of new cases underestimates the real spread of COVID-19 pandemic because of non-tested asymptomatic people and limited global access to reliable diagnostic tests. In this context, COVID-19 mortality with confirmed diagnosis becomes an attractive source of information to be included in the analysis of perspectives and proposals. Objective data are required to calculate the capacity of resources provided by health systems. New strategies are needed to stabilize or minimize the mortality surge. However, we will not afford this goal until more alternatives were available. We still need an effective treatment, an affordable vaccine, or a collective achievement of sufficient immunity (reaching up to 70% of the whole population). At any time, the arriving waves of the pandemic are testing the capacity of governments. The health services struggle to keep the plateau in a steady-state below 100 deaths per million inhabitants. Therefore, it is necessary to increase the alternatives and supplies based on the current and near-future expected demands imposed by the number of deaths by COVID-19. Estimating COVID-19 mortality in various scenarios with the gradual release of social constraints will help predict the magnitude of those arriving waves.

Insurance data can be asymmetric with heavy tails, causing inadequate adjustments of the usually applied models. To deal with this issue, a hierarchical model for collective risk with heavy tails of the claims distributions that also takes into account overdispersion of the number of claims is proposed. In particular, the distribution of the logarithm of the aggregate value of claims is assumed to follow a Student t distribution. Additionally, to incorporate possible overdispersion, the number of claims is modeled as having a negative binomial distribution. Bayesian decision theory is invoked to calculate the fair premium based on the modified absolute deviation utility. An application to a health insurance data set is presented together with some diagnostic measures to identify excess variability. The variability measures are analyzed using the marginal posterior predictive distribution of the premiums according to some competitive models. Finally, a simulation study is carried out to assess the predictive capability of the model and the adequacy of the Bayesian estimation procedure.