Caution over the use of ecological big data for conservation

Evaluating the suitability of large-scale datasets to estimate nitrogen loads and yields across different spatial scales

Decision makers are often confronted with inadequate information to predict nutrient loads and yields in freshwater ecosystems at large spatial scales. We evaluate the potential of using data mapped at large spatial scales (regional to global) and often coarse resolution to predict nitrogen yields at varying smaller scales (e.g., at the catchment and stream reach level). We applied the SPAtially Referenced Regression On Watershed attributes (SPARROW) model in three regions: the Upper Midwest part of the United States, New Zealand, and the Grande River Basin in southeastern Brazil. For each region, we compared predictions of nitrogen delivery between models developed using novel large-scale datasets and those developed using local-scale datasets. Large-scale models tended to underperform the local-scale models in poorly monitored areas. Despite this, large-scale models are well suited to generate hypotheses about relative effects of different nutrient source categories (point and urban, agricultural, native vegetation) and to identify knowledge gaps across spatial scales when data are scarce. Regardless of the spatial resolution of the predictors used in the models, a representative network of water quality monitoring stations is key to improve the performance of large-scale models used to estimate loads and yields. We discuss avenues of research to understand how this large-scale modelling approach can improve decision making for managing catchments at local scales, particularly in data poor regions.

Capability of big data to capture threatened vertebrate diversity in protected areas

Protected areas (PAs) are an essential tool for conservation amid the global biodiversity crisis. Optimizing PAs to represent species at risk of extinction is crucial. Vertebrate representation in PAs is assessed using species distribution databases from the International Union for Conservation of Nature (IUCN) and the Global Biodiversity Information Facility (GBIF). Evaluating and addressing discrepancies and biases in these data sources are vital for effective conservation strategies. Our objective was to gain insights into the potential constraints (e.g., differences and biases) of these global repositories to objectively depict the diversity of threatened vertebrates in the global system of PAs. We assessed differences in species richness (SR) of threatened vertebrates as reported by IUCN and GBIF in PAs globally and then compared how biased this information was with reports from independent sources for a subset of PAs. Both databases showed substantial differences in SR in PAs (t = -62.35, p ≤ 0.001), but differences varied among regions and vertebrate groups. When these results were compared with data from independent assessments, IUCN overestimated SR by 575% on average and GBIF underestimated SR by 63% on average, again with variable results among regions and groups. Our results indicate the need to improve analyses of the representativeness of threatened vertebrates in PAs such that robust and unbiased assessments of PA effectiveness can be conducted. The scientific community and decision makers should consider these regional and taxonomic disparities when using IUCN and GBIF distributional data sources in PA assessment. Overall, supplementing information in these databases could lead to more robust and reliable analyses. Additional efforts to acquire more comprehensive and unbiased data on species distributions to support conservation decisions are clearly needed.

Identifying priority sites for whale shark ship collision management globally

The expansion of the world's merchant fleet poses a great threat to the ocean's biodiversity. Collisions between ships and marine megafauna can have population-level consequences for vulnerable species. The Endangered whale shark (Rhincodon typus) shares a circumglobal distribution with this expanding fleet and tracking of movement pathways has shown that large vessel collisions pose a major threat to the species. However, it is not yet known whether they are also at risk within aggregation sites, where up to 400 individuals can gather to feed on seasonal bursts of planktonic productivity. These "constellation" sites are of significant ecological, socio-economic and cultural value. Here, through expert elicitation, we gathered information from most known constellation sites for this species across the world (>50 constellations and >13,000 individual whale sharks). We defined the spatial boundaries of these sites and their overlap with shipping traffic. Sites were then ranked based on relative levels of potential collision danger posed to whale sharks in the area. Our results showed that researchers and resource managers may underestimate the threat posed by large ship collisions due to a lack of direct evidence, such as injuries or witness accounts, which are available for other, sub-lethal threat categories. We found that constellations in the Arabian Sea and adjacent waters, the Gulf of Mexico, the Gulf of California, and Southeast and East Asia, had the greatest level of collision threat. We also identified 39 sites where peaks in shipping activity coincided with peak seasonal occurrences of whale sharks, sometimes across several months. Simulated collision mitigation options estimated potentially minimal impact to industry, as most whale shark core habitat areas were small. Given the threat posed by vessel collisions, a coordinated, multi-national approach to mitigation is needed within priority whale shark habitats to ensure collision protection for the species.

Machine learning framework for intelligent aeration control in wastewater treatment plants: Automatic feature engineering based on variation sliding layer

Intelligent control of wastewater treatment plants (WWTPs) has the potential to reduce energy consumption and greenhouse gas emissions significantly. Machine learning (ML) provides a promising solution to handle the increasing amount and complexity of generated data. However, relationships between the features of wastewater datasets are generally inconspicuous, which hinders the application of artificial intelligence (AI) in WWTPs intelligent control. In this study, we develop an automatic framework of feature engineering based on variation sliding layer (VSL) to control the air demand precisely. Results demonstrated that using VSL in classic machine learning, deep learning, and ensemble learning could significantly improve the efficiency of aeration intelligent control in WWTPs. Bayesian regression and ensemble learning achieved the highest accuracy for predicting air demand. The developed models with VSL-ML models were also successfully implemented under the full-scale wastewater treatment plant, showing a 16.12 % reduction in demand compared to conventional aeration control of preset dissolved oxygen (DO) and feedback to the blower. The VSL-ML models showed great potential to be applied for the precision air demand prediction and control. The package as a tripartite library of Python is called wwtpai, which is freely accessible on GitHub and CSDN to remove technical barriers to the application of AI technology in WWTPs.