NOAA National Marine Fisheries Service

Environmental DNA (eDNA) metabarcoding can rapidly characterise biodiversity, yet its accuracy and effectiveness are limited by incomplete DNA barcode reference databases. We evaluated how comprehensive reference databases that include sequence variation within genomes (intragenomic) and across individuals and species (intergenomic) improve eDNA-based biodiversity assessments. We collected coral tissue and water samples at deep sites offshore Puerto Rico for reference barcoding and eDNA metabarcoding. Genome skimming coral specimens yielded 28S barcodes for 314 of 346 samples (90.8%) and revealed divergent intragenomic 28S lineages in multiple octocoral families. Incorporating local reference barcodes substantially changed ASV taxonomic classifications: 22 ASVs (8.9%) gained genus-level resolution, 19 ASVs (7.7%) were reassigned to different genera, and 14 ASVs (5.7%) lost incorrect genus-level classifications. Thus, incomplete reference databases produce not only unclassified ASVs but also false positive detections and ecologically meaningful misclassifications. When intragenomic 28S lineages were excluded from the reference database, 18 ASVs (7.4%) could not be classified to family or genus, demonstrating that unrecognised intragenomic variation can be mistaken for unsampled taxa. Integrating reference genome skimming and eDNA metabarcoding expanded known coral family richness by 36% at depths shallower than 1000 m and by 181% at depths greater than 1000 m. eDNA also detected two coral families previously unknown off Puerto Rico and nearby islands, underscoring its potential for biodiversity discovery.

Accurate estimation of daily reference evapotranspiration (ETo) is essential for effective irrigation scheduling, improved water-use efficiency, and sustainable crop production in arid and semi-arid regions. Although the FAO-56 Penman–Monteith equation (ETo PM ) is widely accepted as the reference method, its reliance on complete meteorological data limits its applicability in data-scarce agricultural systems such as those in eastern Morocco. This study proposes a practical solution for daily ETo estimation under data-limited conditions by evaluating machine learning approaches against traditional empirical models. Daily climatic data (2001–2025) were collected from airport meteorological stations, NASA POWER reanalysis, and on-farm sensors at four representative locations in Eastern Morocco (Oujda, Berkane, Taourirt, and Figuig). ETo was computed using the ETo PM equation. Three ML algorithms—Support Vector Regression (SVR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost)—were developed under multiple input scenarios guided by feature-importance analysis, which identified solar radiation (Rsn), maximum temperature (Tmax), and relative humidity (RH) as dominant predictors. To ensure realistic evaluation and avoid information leakage, a strict chronological split was applied, using 2001–2018 for training and 2019–2025 for independent testing. The best ML configurations were then compared with widely used empirical ETo equations (Hargreaves–Samani, Jensen–Haise, Priestley–Taylor, Makkink, and Turc). Results showed that XGBoost consistently outperformed SVR and RF, achieving the best balance between accuracy and computational efficiency. The three-input configuration (Rsn + Tmax + RH) produced near-reference performance across all locations ( R 2 = 0.976; RMSE < 0.55 mm day −1 ; training time ≈ 4.48 s; RAM ≈ 0.13 MB), while the reduced two-input configuration (Rsn + Tmax) maintained reliable performance ( R 2 = 0.936; RMSE ≈ 0.60 mm day −1 ; training time ≈ 0.33 s; RAM ≈ 0.13 MB). In contrast, empirical approaches exhibited poor predictive capability and weak transferability, with negative R 2 values and substantially larger errors (RMSE generally > 4 mm day −1 , exceeding 7 mm day −1 in the weakest formulations). Overall, this machine learning–based framework provides a robust and operational alternative for daily ETo estimation in eastern Morocco, supporting efficient irrigation scheduling, improved agricultural water management, enhanced crop productivity, and sustainable agriculture in arid and semi-arid regions.

BackgroundJuvenile loggerhead sea turtles represent a crucial but understudied stage of the species life cycle. Most studies have been conducted at regional scales and there remains a need for a broader scale synthesis. In the North Atlantic, loggerheads encounter numerous threats such as fishing during their transoceanic journey, so it is important to get a broad view of their distribution, movement and habitat preferences.MethodsFor this study, we gathered tracking data from 124 juvenile loggerheads tagged along the Azores, Canary Islands, Madeira and in the western North Atlantic to develop a habitat model for the entire basin. We used a SSM to interpolate tracks to daily positions, simulated pseudo-absences with a correlated random walk for background data, and explored 17 environmental variables using boosted regression trees to select the best model in terms of biological realism and predictive power. The best model obtained (AUC = 0.988) was then used to predict habitat suitability for the North Atlantic for the period 1998–2022. Stranding records of loggerheads in France, UK and Ireland were used to discuss the predictions.ResultsAfter filtering, tracks of 105 individuals, covering most of the basin, were used in the model. Our predictions of habitat suitability show the importance of the Gulf Stream and the Azorean Current. Habitat suitability is predicted to be higher in the area between 30° and 45°N all year round, while open waters below 30°N, including Cape Verde, are not predicted to be a suitable habitat. In northern and eastern areas, like the Bay of Biscay and UK and Irish waters, habitat suitability varied seasonally, with colder seasons showing lower values. Accordingly, strandings in those areas occurred mostly during winter (December-February; 45.1%) and spring (March-May; 34.1%).ConclusionsThe model offers a first basin-scale prediction of the seasonal distribution of juvenile loggerheads in the North-Atlantic, which shows consistency with stranding and bycatch data in the basin. This study represents a first step towards a broader scale understanding of juveniles’ habitat preferences that can be used to quantify the magnitude and extent of the threats that they face.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40462-026-00640-2.

We evaluated 13 years of fishery-independent small juvenile (< 1.86 m stretch total length) endangered smalltooth sawfish sampling data from southwest Florida’s Charlotte Harbor estuarine system. From 2010 through 2022, within the sampling domain, estimated abundance declined, with an annual average across all seasons of 34 ± 11 individuals (range: 15–69). High-use areas were spatiotemporally stable and associated with well-oxygenated (dissolved oxygen > 7 mg/L), warm (25–33 °C), and brackish (salinity 5–27) waters between 5 and 15 km upriver along natural shorelines with concentrations of mangroves in both riverine nurseries. Adult female abundance, back-calculated from expected brood size (48 ± 16, range: 26–144) and stable age distribution (87 ± 30, range: 39–211), was low. Our findings of a significant declining trend in relative abundance of small juveniles and limited numbers of adult females are concerning considering ongoing threats from commercial shrimp trawl bycatch, destruction of nursery habitats, and major recent mortality events. Our model facilitates estimation of the number of individuals potentially exposed to harmful anthropogenic activities and identifies perennial high-use areas as priorities for permanent protection from habitat alterations and loss to promote recovery of the endangered smalltooth sawfish population in this unique estuarine system.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-14430-0.