Explore the vast range of titles available.

Ecological niche theory predicts sympatric species to show segregation in their spatio‐temporal habitat utilization or diet as a strategy to avoid competition. Similarly, within species individuals may specialize on specific dietary resources or foraging habitats. Such individual specialization seems to occur particularly in environments with predictable resource distribution and limited environmental variability. Still, little is known about how seasonal environmental variability affects segregation of resources within species and between closely related sympatric species. The aim of the study was to investigate the foraging behaviour of three closely related and sympatrically breeding fulmarine petrels (Antarctic petrels Thalassoica antarctica, cape petrels Daption capense and southern fulmars Fulmarus glacialoides) in a seasonally highly variable environment (Prydz Bay, Antarctica) with the aim of assessing inter‐ and intraspecific overlap in utilized habitat, timing of foraging and diet and to identify foraging habitat preferences. We used GPS loggers with wet/dry sensors to assess spatial habitat utilization over the entire breeding season. Trophic overlap was investigated using stable isotope analysis based on blood, feathers and egg membranes. Foraging locations were identified using wet/dry data recorded by the GPS loggers and expectation‐maximization binary clustering. Foraging habitat preferences were modelled using generalized additive models and model cross‐validation. During incubation and chick‐rearing, the utilization distribution of all three species overlapped significantly and species also overlapped in the timing of foraging during the day—partly during incubation and completely during chick‐rearing. Isotopic centroids showed no significant segregation between at least two species for feathers and egg membranes, and among all species during incubation (reflected by blood). Within species, there was no individual specialization in foraging sites or environmental space. Furthermore, no single environmental covariate predicted foraging activity along trip trajectories. Instead, best‐explanatory environmental covariates varied within and between individuals even across short temporal scales, reflecting a highly generalist behaviour of birds. Our results may be explained by optimal foraging theory. In the highly productive but spatio‐temporally variable Antarctic environment, being a generalist may be key to finding mobile prey—even though this increases the potential for competition within and among sympatric species. This article shows a striking case of closely related and sympatrically breeding species overlapping in their resource use across multiple dimensions. The very unusual results are best explained by high food availability in combination with spatio‐temporal environmental variability and mobile prey fields, requiring predators to respond opportunistically and show generalist foraging behaviour.

We present the results of a validation of atmospheric inversions of CO2 fluxes using four transport models. Each inversion uses data primarily from surface stations, combined with an atmospheric transport model, to estimate surface fluxes. The validation (or model evaluation) consists of running these optimized fluxes through the forward model and comparing the simulated concentrations with airborne concentration measurements. We focus on profiles from Cape Grim, Tasmania, and Carr, Colorado, while using other profile sites to test the generality of the comparison. Fits to the profiles are generally worse than to the surface data from the inversions and worse than the expected model‐data mismatch. Thus inversion estimates are generally not consistent with the profile measurements. The TM3 model does better by some measures than the other three models. Models perform better over Tasmania than Colorado, and other profile sites bear out a general improvement from north to south and from continental to marine locations. There are also errors in the interannual variability of the fit, consistent in time and common across models. This suggests real variations in sources visible to the profile but not the surface measurements. Key Points Atmospheric CO2 inversion multimodel comparison Inversion validation using aircraft data

The coronavirus disease 2019 (COVID-19) pandemic had brought negative consequences and new stressors to mothers. The current study aims to compare factors predicting maternal mental health during the COVID-19 lockdown in China, Italy, and the Netherlands. The sample consisted of 900 Dutch, 641 Italian, and 922 Chinese mothers (age M = 36.74, s.d. = 5.58) who completed an online questionnaire during the lockdown. Ten-fold cross-validation models were applied to explore the predictive performance of related factors for maternal mental health, and also to test similarities and differences between the countries. COVID-19-related stress and family conflict are risk factors and resilience is a protective factor in association with maternal mental health in each country. Despite these shared factors, unique best models were identified for each of the three countries. In Italy, maternal age and poor physical health were related to more mental health symptoms, while in the Netherlands maternal high education and unemployment were associated with mental health symptoms. In China, having more than one child, being married, and grandparental support for mothers were important protective factors lowering the risk for mental health symptoms. Moreover, high SES (mother's high education, high family income) and poor physical health were found to relate to high levels of mental health symptoms among Chinese mothers. These findings are important for the identification of at-risk mothers and the development of mental health promotion programs during COVID-19 and future pandemics.

Installing rooftop solar photovoltaic (PV) with an on-grid system benefits consumers because it can reduce imports of electrical energy from the grid. This study aims to model the estimation of energy imports generated from on-grid rooftop solar PV systems. This estimation model was carried out in 20 provincial capitals in Indonesia. The parameters used are weather conditions, orientation angle, and energy generated from the on-grid rooftop solar PV system. The value of imported energy is divided into three combinations based on the azimuth angle direction, which describes the type and shape of the roof of the building (one-direction, two-directions, and three-directions). Modeling was done using machine learning with neural network (NN), linear regression, and support vector machine. A comparison of the machine learning algorithm results is NN produces the smallest root mean square error (RMSE) value of the three. Model enhancement uses a grid search cross-validation (GSCV) to become the GSCV-NN model. The RMSE results were enhanced from 53.184 to 44.389 in the one-direction combination, 145.562 to 141.286 in the two-direction combination, and 81.442 to 76.313 in the three-direction combination. The imported energy estimation model on the on-grid rooftop solar PV system with GSCV-NN produces a more optimal and accurate model.

Background and Aim: The increasing adoption of artificial intelligence (AD in medical research offered alternative methods for medical data processing. This study evaluated comparatively the predictive performance of feedforward neural networks (FFNN) regression versus classical statistical regression analysis in estimating the risk of post-COVID-19 type 2 diabetes based on metabolic factors. The primary objective was to assess the applicability, advantages, and limitations of these approaches when applied to relatively small medical datasets. Materials and Methods: We started with the analysis of a small dataset - 130 patient records with metabolic parameters [1]. The risk of post-COVID-19 type 2 diabetes (glycaemia at 4 and at 12 months post-COVID as function of metabolic parameters) was predicted using both linear regression and FFNN. The regression model followed standard statistical guidelines, while the FFNN required optimization of hyperparameters, including the number of layers, activation functions, learning rate, and optimization algorithms. We extended the study using simulated data to further compare logistic regression (a data set of 300 patients) with neural networks. Results: The classical regression models demonstrated stable performance with clear interpretability, offering well-defined coefficients and statistical significance measures. However, FFNN did not yield superior predictive accuracy, and its performance varied significantly depending on the choice of hyperparameters. The optimization process for NN required extensive trial and error, as no universal guidelines for parameter selection were applicable in this context. Discussion. Our findings highlight a real challenge in medical AI applications for data processing: when dealing with small datasets, neural networks do not necessarily outperform classical methods. Regression provided robust results with minimal computational effort, while FFNN required complex tuning without a clear performance advantage. The use of simulated data revealed that NN might be more effective in larger datasets with potential non-linear patterns, but limited interpretability. Conclusion: Al-based models are, indeed, recommended for data processing of large and/or unstructured complex medical data sets. However, as a conclusion of this study, regression models proved to be a more practical and reliable choice for small-scale medical predictions. Future work should explore hybrid models that combine interpretability with non-linear modeling capacity to optimize predictive accuracy in clinical settings.

The farmer's income and crop yield are significantly affected in Punjab, Pakistan due to poor-quality groundwater. To observe, monitor and categorized groundwater quality, this research study was carried out in Faisalabad (FSD) and Toba Tek Singh (TTS) districts of Punjab, Pakistan to check its suitability for irrigation with three major parameters (i.e. EC, SAR, and RSC). Geo-statistical water quality analysis was carried out using the GS+ and ArcGIS includes three following basic components: normalized histograms, semivariograph, and Kriging. The cross-validation techniques were used to determine the accuracy. A hydro-economic model was applied to observe the impact of groundwater quality on crop yield and farmers’ income. It was found that the percent area under a good groundwater quality zone in FSD was about 25% fewer than TTS. In FSD, the majority area of the aquifer was under marginal (50–55%) to poor (39–44%) quality groundwater zones and salinity and sodicity are major threats depicted by EC and RSC, respectively. In TTS district, salinity was the only major risk to groundwater quality as about 45% area was under poor quality zone. The overall aquifer’s area under about good (~ 33%), marginal (~ 29%) and poor (~ 38%) quality groundwater zone. It was found that the impact of the monsoon season was found not considerable on the groundwater quality of both districts. Comparing the economic models in two districts using the different quality water it was found that the BCR (Benefit Cost Ratio) was recorded 2.31, 2.13 and 1.73 in FSD district while in TTS district the BCR was 2.35, 2.09 and 1.58 for good, marginal and poor-quality zone, respectively. The results of the research recommend that monitoring and mapping of groundwater are necessary for proper management of groundwater resources, leading to reduced economic losses and increased crop yield.

Hierarchical models defined by means of directed, acyclic graphs are a powerful and widely used tool for Bayesian analysis of problems of varying degrees of complexity. A simulation-based method for model criticism in such models has been suggested by O'Hagan in the form of a conflict measure based on contrasting separate local information sources about each node in the graph. This measure is however not well calibrated. In order to rectify this, alternative mutually similar tail probability-based measures have been proposed independently and have been proved to be uniformly distributed under the assumed model in quite general normal models with known covariance matrices. In the present paper, we extend this result to a variety of models. An advantage of this is that computationally costly pre-calibration schemes needed for some other suggested methods can be avoided. Another advantage is that non-informative prior distributions can be used when performing model criticism.

Six non-linear growth functions were fitted to tree height–diameter data of ten conifer species collected in the inland Northwest of the United States. The data sets represented a wide range of tree sizes, especially large-sized trees. According to the model statistics, the six growth functions fitted the data equally well, but resulted in different asymptote estimates. The model prediction performance was evaluated using Monte Carlo cross-validation or data splitting for 25-cm diameter classes. All six growth functions yielded similar mean prediction errors for small- and middle-sized trees. For large-sized trees [e.g. DBH (diameter at breast height)>100 cm], however, five of the six growth functions (except the Gompertz function) overestimated tree heights for western white pine, western larch, Douglas-fir, subalpine fir, and ponderosa pine, but underestimated tree heights for western hemlock and Engelmann spruce. Among these five functions, the Korf/Lundqvist and Exponential functions produced larger overestimations. The Schnute, Weibull, and Richards functions were superior in prediction performance to others. The Gompertz function seemed always to underestimate tree heights for large-sized trees.

Abstract It is known that model averaging estimators are useful when there is uncertainty governing which covariates should enter the model. We argue that in applied research there is also uncertainty as to which method one should deploy, prompting model averaging over user-defined choices. Specifically, we propose, and detail, a nonparametric regression estimator averaged over choice of kernel, bandwidth selection mechanism and local-polynomial order. Simulations and an empirical application are provided to highlight the potential benefits of the method.