Gustavo Fonseca

Biodiversity monitoring programs need to deliver accurate, timely, and actionable predictions. To establish a predictive monitoring program for deep‐sea benthos of the Santos Basin, Brazil, we developed a two‐stage structured model that allowed comparison of biodiversity predictions obtained from environmental simulations (2M‐Sim). We also modeled the environmental variables as a function of spatial and temporal variables and compared this model's predictions with predictions obtained from real environmental data (2M). We built unstructured models (1M) as references to evaluate whether the proposed structured approach was reliable. We expected no significant differences between 1M and 2M or between 2M and 2M‐Sim. Data were obtained from 100 stations at depths of 25–2400 m during two surveys (2019 and 2021). In our model framework, we used 12 benthic macro‐ and meiofaunal variables, 44 sediment and water column environmental variables, and four spatial and temporal variables. We applied a random forest algorithm to the structured and unstructured models. All comparisons were performed with 20% of the dataset set aside for validation. The average accuracy was 72%, 69%, and 68% for the 1M, 2M, and 2M‐Sim models, respectively. Accuracies of 2M ranged from 38% to 84% and were generally higher for macrofauna. The observed accuracy loss from 1M to 2M (3%) and from 2M to 2M‐Sim (1%) was not significant for any biodiversity variable. The 2M model identified 30 significant environmental variables; bottom water parameters and sedimentary phytopigment and carbonate concentrations were the best predictors. Our approach supports biodiversity conservation by optimizing data needs and future sampling and by guiding data‐driven management decisions for benthic biodiversity.

Environmental exposure to heavy metals has been associated with cancer. Nonetheless, the effects of exposure to combined heavy metals on cancer mortality of urinary and gastrointestinal systems have not been examined. This study assessed the long-term association between environmental exposure to multiple heavy metals and cancer mortality rates in the digestive and urinary tracts. The death rates/100,000 inhabitants over 41 years were collected from Santo Amaro, the exposed town with a former lead smelter, and Ribeira do Pombal, the reference town, in Bahia, Brazil. The exposure duration, age and gender were considered predictors. The overall accuracy of the models ranged from 84 to 97%. For the population above 80 years, the probability of total mortality went from 20% in the first 6 years of exposure to over 70% after 20 years. The same trend was observed for the urinary system. For the digestive system, the mortality in the elders started after 15 years of exposure and peaked with a 30% probability after 35 years. In conclusion, environmental exposure to multiple heavy metals was linked to cancer-related deaths in the gastrointestinal and urinary tracts of the elderly population.

This study aimed to apply a machine learning workflow to identify the most relevant biological parameters for predicting both the dose and the route of crack cocaine administration.Seventeen variables were evaluated in rats exposed to different doses of crack cocaine, either intraperitoneally (18 or 36 mg/kg, i.p.) or via passive inhalation (25, 50, or 100 mg). Random forest analysis was used to build predictive models, feature importance analysis to identify key variables, and interaction dependence analysis to explore relationships among variables.Eighty percent of the data was used for training and twenty percent for testing. The model achieved 85% accuracy in the training phase and 100% in the test phase. During training, the highest accuracy was observed for the 100 mg inhaled group, while the lowest was for the 50 mg inhaled group. Notably, 20% of the 50 mg i.p. cases were misclassified as 36 mg i.p. Feature importance analysis highlighted four key predictors: liver karyolysis, kidney Ki-67 expression, liver binucleation, and escape behavior.These findings demonstrate that machine learning can accurately predict both dose and route of crack cocaine exposure, and can highlight biologically relevant parameters involved in the drug's systemic and behavioral effects.

This paper identifies the top-50 priority questions for meiofaunal research, highlighting their critical roles in biogeochemical cycles and biodiversity. It calls for a balanced research agenda, international cooperation, and advances in technology to overcome current challenges and unlock meiofauna’s full potential. This paper identifies the top-50 priority questions for meiofaunal research, highlighting their critical roles in biogeochemical cycles and biodiversity. It calls for a balanced research agenda, international cooperation, and advances in technology to overcome current challenges and unlock meiofauna’s full potential.

Microbial communities in marine ecosystems play a pivotal role in global biogeochemical cycles, with particular relevance in ecologically and industrially significant regions such as the Santos Basin (SB), Brazil’s largest marine sedimentary basin and a hub for oil and gas exploration. Yet, our capacity of predicting the microbial community structure and function remains limited for marine ecosystems. This study investigated the structure of microbial communities across different depths in the SB, using amplicon sequencing data and quantitative cell counts obtained via flow cytometry. Using a hybrid machine learning framework combining Self-Organizing Maps and Random Forest, we identified five distinct microbial assemblages (named microbial associations) in the SB predicted with 86% accuracy. These associations were primarily driven by temperature, water density, salinity, and nutrients such as phosphate and nitrate. Our findings showed a clear stratification of microbial communities across pelagic zones, with temperature as the main factor driving the structure in epipelagic and mesopelagic layers, while salinity and density exerted greater influence in the deeper bathypelagic communities. Temporal and spatial variations, particularly between 2019 and 2021, in regions influenced by the Cabo Frio upwelling and Rio de la Plata plume, further highlighted the impact of regional and local oceanographic processes on community dynamics. The associations from deeper waters harbored more diverse microbial assemblages, and shallow waters, on the other hand, possess higher absolute abundance of microbial cells, suggesting niche specialization across depths. This study underscores the importance of environmental gradients as well as local oceanographic processes in shaping microbial diversity, providing valuable insights into the ecological dynamics of the SB, which are essential for understanding the potential impacts of human activities, such as oil and gas exploration and production, on these critical marine ecosystems.

Meiofauna—a collective term to define microscopic animals—represent a numerically important component of biodiversity in most of Earth’s ecosystems and play a crucial role in biogeochemical cycles. Meiofauna have also been used as models to understand fundamental adaptive processes, have contributed to a better understanding of the animal’s Tree of Life, and are believed to be a treasure trove for future genomic studies. To celebrate the diversity of research topics brought to us by the term “meiofauna”, we gathered a multidisciplinary team of 42 ecologists, taxonomists, morphologists, biogeographers, molecular biologists, and scientific disseminators to list 194 fundamental questions in meiofaunal research. Then, through an online survey, 251 scientists, administrators, students, and stakeholders assisted us in reducing this list to 50 top-priority questions. Applied topics related to anthropogenic impact and climate change received the highest scores, whereas questions related to areas in development such as genomics or adaptations, received less attention. Whereas we might not be exploiting meiofauna’s full potential yet, more and more integrative approaches and technological developments will create opportunities to employ these fascinating organisms to answer broad and important questions, despite of their impediments related to their small body size. Meiofauna research agenda should balance amongst investigating general questions, addressing more specialized research topics, and generating primary data on distribution, taxonomy, traits, and DNA sequences. The geographical and taxonomic biases that have historically affected meiofaunal research can be alleviated by promoting international cooperation, open data sharing, and an increase effort in education, taxonomic training, as well as scientific communication. We hope that this will get both researchers and the general public intrigued by those small critters that constantly lurk unseen in front of us.

This special issue examines the latest discoveries from the interdisciplinary project, “Regional Characterization of the Santos Basin – (PCR-BS),” designed by the Brazilian oil company Petrobras as part of its environmental program initiatives. Future environmental management practices are threatened by lack of comprehensive oceanographic knowledge in offshore oil & gas regions. This concern inspired this special issue. The 13 articles here present data primarily gathered during the 2019 campaigns. The study collected physical, chemical, sedimentological, and biological data from both benthic and pelagic environments in the basin. The findings are classified into four categories: methodological aspects, hydrographic conditions, and pelagic and benthic ecosystems. The scientific research presented in this special issue represents an initial effort towards reconciling sustainable oil and gas exploration and production with the preservation of regional biodiversity and environmental health.

This study aims model the distribution of meiofauna indicators in relation to environmental variables from the Santos Basin continental margin, SE Brazil, using machine learning techniques, to provide baseline information and foster future monitoring programs. A total of 100 sampling stations were distributed in eight transects and 11 isobaths (25 to 2,400 m) perpendicular to the coast. In each station, three replicates were sampled for meiofauna and 38 environmental parameters. A total of 28 meiofauna taxa were found, with a mean richness varying from 3 to 15 taxa per station. Meiofauna mean density varied between 55 and 2,001 ind. 10 cm-2. Density of meiofauna and its most frequent taxa (Nematoda, Copepoda, Kinorhyncha, and Polychaeta), and taxa richness were used as descriptors for the models. Meiofauna and nematode density showed the highest training and testing accuracies, with R² values above 0.74. Based on the distribution of meiofauna descriptors and their responses to environmental conditions, we suggest a mosaic of six benthic zones. The La Plata Plume zone and the Cabo Frio Upwelling zone are two of the most diverse and productive zones in the continental shelf, wich are separated by the less productive Central Continental Shelf zone. A fourth zone, with very low meiofauna densities, corresponds to the carbonated sediments of the shelf-break. The Upper and Mid-Slope is a narrow zone along the entire basin, with intermediate densities and small amounts of high-quality organic carbon. The largest, impoverished zone, the Lower Slope and Plateau comprises the deepest areas and the São Paulo Plateau. The study showed that, although some zones can be recognized by most meiofauna descriptors, others are better characterized by specific ones, implying that meiofauna indicators should be monitored concomitantly. We recommend the optimization of sampling design based on our model to reduce costs and increase our understanding of the system.

Integrating intricate environmental data within a unified analytical framework for extensive conservation and monitoring initiatives encounters several challenges. These challenges encompass defining a conceptual model outlining cause-and-effect relationships, addressing dissimilarities in data source quantity and information content, grappling with missing or noisy data, fine-tuning model optimization, achieving accurate predictions, and tackling the issue of imbalanced observations across factors. In the context of the Santos project, dedicated to comprehending the spatio-temporal dynamics of benthic, pelagic, and physical systems for the facilitation of conservation and monitoring programs, the application of machine learning’s random forest (RF) technique for modeling univariate data offers notable advantages. This approach adeptly handles non-linearity, covariation, and interactive effects among predictors. For modeling multivariate data sets, a hybrid strategy combining a self-organizing map (SOM) and RF is harnessed to effectively tackle the challenges. Addressing missing values, the bagging imputation technique demonstrated superior performance compared to other methods. Both machine learning techniques discussed herein exhibit resilience against the impact of noisy data, yet the identification of noisy data remains feasible based on model outputs. In scenarios of imbalanced data sets, we investigate the correlation between the RF model’s overall statistics and those of individual classes. The joint interpretation of these statistics aids in comprehending model limitations and facilitates discussions on the environmental mechanisms shaping observed patterns. We propose two analytical workflows that not only enable the exploration and enhancement of model accuracy but also facilitate the investigation of potential cause-and-effect relationships inherent in the data. Furthermore, these workflows lay the foundation for implementing long-term learning algorithms, a pivotal increment for monitoring initiatives. Notably, these workflows, alongside the discussed analytical challenges, can be seamlessly implemented within iMESc, an open-source application.