Explore the vast range of titles available.

Aim: To synthesize the species distribution modelling (SDM) literature to inform which variables have been used in MaxEnt models for different taxa and to quantify how frequently they have been important for species' distributions. Location: Global. Methods: We conducted a quantitative synthesis analysing the contribution of over 400 distinct environmental variables to 2040 MaxEnt SDMs for nearly 1900 species representing over 300 families. Environmental variables were grouped into 24 related factors and results were analysed by examining the frequency with which variables were found to be most important, the mean contribution of each variable (at various taxonomic levels), and using TrueSkill™, a Bayesian skill rating system. Results: Precipitation, temperature, bathymetry, distance to water and habitat patch characteristics were the most important variables overall. Precipitation and temperature were analysed most frequently and one of these variables was often the most important predictor in the model (nearly 80% of models, when tested). Notably, distance to water was the most important variable in the highest proportion of models in which it was tested (42% of 225 models). For terrestrial species, precipitation, temperature and distance to water had the highest overall contributions, whereas for aquatic species, bathymetry, precipitation and temperature were most important. Main conclusions: Over all MaxEnt models published, the ability to discriminate occurrence from reference sites was high (average AUC = 0.92). Much of this discriminatory ability was due to temperature and precipitation variables. Further, variability (temperature) and extremes (minimum precipitation) were the most predictive. More generally, the most commonly tested variables were not always the most predictive, with, for instance, 'distance to water' infrequently tested, but found to be very important when it was. Thus, the results from this study summarize the MaxEnt SDM literature, and can aid in variable selection by identifying underutilized, but potentially important variables, which could be incorporated in future modelling efforts.

Maxent, one of the most commonly used methods for inferring species distributions and environmental tolerances from occurrence data, allows users to fit models of arbitrary complexity. Model complexity is typically constrained via a process known as L 1 regularization, but at present little guidance is available for setting the appropriate level of regularization, and the effects of inappropriately complex or simple models are largely unknown. In this study, we demonstrate the use of information criterion approaches to setting regularization in Maxent, and we compare models selected using information criteria to models selected using other criteria that are common in the literature. We evaluate model performance using occurrence data generated from a known \"“true\"” initial Maxent model, using several different metrics for model quality and transferability. We demonstrate that models that are inappropriately complex or inappropriately simple show reduced ability to infer habitat quality, reduced ability to infer the relative importance of variables in constraining species' distributions, and reduced transferability to other time periods. We also demonstrate that information criteria may offer significant advantages over the methods commonly used in the literature.

Conventional machine-learning (ML) models in computational chemistry learn to directly predict molecular properties using quantum chemistry only for reference data. While these heuristic ML methods show quantum-level accuracy with speeds several orders of magnitude faster than traditional quantum chemistry methods, they suffer from poor extensibility and transferability; i.e., their accuracy degrades on large or new chemical systems. Incorporating quantum chemistry frameworks into the ML models directly solves this problem. Here we take the structure of semiempirical quantum mechanics (SEQM) methods to construct dynamically responsive Hamiltonians. SEQM methods use empirical parameters fitted to experimental properties to construct reduced-order Hamiltonians, facilitating much faster calculations than ab initio methods but with compromised accuracy. By replacing these static parameters with machine-learned dynamic values inferred from the local environment, we greatly improve the accuracy of the SEQM methods. Trained on molecular energies and atomic forces, these dynamically generated Hamiltonian parameters show a strong correlation with atomic hybridization and bonding. Trained with only about 60,000 small organic molecular conformers, the resulting model retains interpretability, extensibility, and transferability when testing on much larger chemical systems and predicting various molecular properties. Overall, this work demonstrates the virtues of incorporating physics-based descriptions with ML to develop models that are simultaneously accurate, transferable, and interpretable.

Aim MaxEnt has been widely used to model species’ geographic distributions as functions of environmental variables and to predict changes in distributions in response to environmental change. Here, we test the predictive ability of MaxEnt models through time by modelling colonizations and extinctions. Location North America. Methods Using data for 21 species from the North American Breeding Bird Survey, we first related avian species’ geographic distributions to the spatial variation in environmental conditions. Then, we modelled site‐specific colonizations and extinctions between 1979 and 2009 as functions of MaxEnt‐estimated habitat suitability and neighbourhood occupancy. Results We found that colonization and extinction probabilities were related to spatial variation in habitat suitability, and to neighbourhood occupancy, in the expected directions. However, change in habitat suitability (which is much smaller through time than through space) is a weak predictor of extinction and worse for colonization. This is because a) for most species and most sites, climatic suitability did not change dramatically between 1979 and 2009, and b) the relationship between colonization or extinction probability and change in climatic variables is very weak (r2 = 0.02). Most colonizations and extinctions are apparently unrelated to climate change. Main conclusions MaxEnt models apparently capture a real effect of habitat suitability on North American bird species’ distributions, but over short and medium time scales, occupancy of neighbouring sites by conspecifics predicts changes in occupancy as well as, or better than changes in climatic habitat suitability, as characterized by MaxEnt. One would not expect species’ distributions to track climate change closely. Prediction of species’ responses to climate change should 1) recognize that the process of colonization and extinction are not equally well predicted by species distribution models and 2) account for the spatial structure of species’ distributions.

Leaf hyperspectral reflectance (HSR) data have gained increasing attention due to their usage in predicting a range of leaf physiological, biochemical, structural, and photosynthetic traits using machine learning (ML) models. The PROSPECT family of models offers a complementary, mechanistic means to estimate leaf traits from HSR data using model inversion. However, a comprehensive evaluation of the accuracy and transferability of the PROSPECT model across a large set of species is hindered by the limited availability of ground truth data sets. Here, we employed a combination of inversion and forward simulation of the PROSPECT-D model across a broad range of species and identified four narrow wavebands linked to environmental effects. We also introduced a novel framework using partial least squares regression to enable the analysis of the transferability of the machine learning models trained base on the PROSPECT-D across species. This analysis revealed trait-specific patterns of transferability for the machine learning surrogate based on the PROSPECT-D forward model. We then extended this analysis to PROSPECT-D inversion using neural networks and developed a fast, accurate deep-learning-based surrogate inversion approach to estimate leaf traits from measured HSR data. Our data-driven framework paves the way for improving the accuracy of PROSPECT and similar mechanistic models.

Bioclimatic envelope models are commonly used to assess the influence of climate change on species' distributions and biodiversity patterns. Understanding how methodological choices influence these models is critical for a comprehensive evaluation of the estimated impacts. Here we systematically assess the performance of bioclimatic envelope models in relation to the selection of predictors, modeling technique, and pseudo‐absences. We considered (a) five different predictor sets, (b) seven commonly used modeling techniques and an ensemble model, and (c) three sets of pseudo‐absences (1,000 pseudo‐absences, 10,000 pseudo‐absences, and the same as the number of presences). For each combination of predictor set, modeling technique, and pseudo‐absence set, we fitted bioclimatic envelope models for 300 species of mammals, amphibians, and freshwater fish, and evaluated the predictive performance of the models using the true skill statistic (TSS), based on a spatially independent test set as well as cross‐validation. On average across the species, model performance was mostly influenced by the choice of predictor set, followed by the choice of modeling technique. The number of the pseudo‐absences did not have a strong effect on the model performance. Based on spatially independent testing, ensemble models based on species‐specific nonredundant predictor sets revealed the highest predictive performance. In contrast, the Random Forest technique yielded the highest model performance in cross‐validation but had the largest decrease in model performance when transferred to a different spatial context, thus highlighting the need for spatially independent model evaluation. We recommend building bioclimatic envelope models according to an ensemble modeling approach based on a nonredundant set of bioclimatic predictors, preferably selected for each modeled species. Bioclimatic envelope models are commonly used to assess the influence of climate change on species' distributions and biodiversity patterns. We systematically evaluated the performance of bioclimatic envelope models in relation to the selection of predictors, modelling technique, and pseudo‐absences. Model performance was mostly influenced by the choice of predictor set, followed by the choice of modelling technique, and based on our results, we recommend building bioclimatic envelope models according to an ensemble modelling approach based on a nonredundant set of bioclimatic predictors, preferably selected for each modelled species.

Habitat selection models frequently use data collected from a small geographic area over a short window of time to extrapolate patterns of relative abundance into unobserved areas or periods of time. However, such models often poorly predict the distribution of animal space‐use intensity beyond the place and time of data collection, presumably because space‐use behaviors vary between individuals and environmental contexts. Similarly, ecological inference based on habitat selection models could be muddied or biased due to unaccounted individual and context dependencies. Here, we present a modeling workflow designed to allow transparent variance‐decomposition of habitat‐selection patterns, and consequently improved inferential and predictive capacities. Using global positioning system (GPS) data collected from 238 individual pronghorn, Antilocapra americana, across three years in Utah, USA, we combine individual‐year‐season‐specific exponential habitat‐selection models with weighted mixed‐effects regressions to both draw inference about the drivers of habitat selection and predict space‐use in areas/times where/when pronghorn were not monitored. We found a tremendous amount of variation in both the magnitude and direction of habitat selection behavior across seasons, but also across individuals, geographic regions, and years. We were able to attribute portions of this variation to season, movement strategy, sex, and regional variability in resources, conditions, and risks. We were also able to partition residual variation into inter‐ and intra‐individual components. We then used the results to predict population‐level, spatially and temporally dynamic, habitat‐selection coefficients across Utah, resulting in a temporally dynamic map of pronghorn distribution at a 30 × 30 m resolution but an extent of 220 000 km2. We believe our transferable workflow can provide managers and researchers alike a way to turn limitations of traditional habitat selection models – variability in habitat selection – into a tool to understand and predict species‐habitat associations across space and time.

Transferability is key to many of the most novel and interesting applications of ecological niche models, such that maximizing predictive power of model transfers is crucial. Here, we explored consensus methods as a means of reducing uncertainty and improving model transferability in anticipating the potential distribution of an invasive moth ( Hyphantria cunea ). Individual native-range niche models were calibrated using seven modelling algorithms and four environmental datasets, representing different degrees of dimensionality, spatial correlation, and ecological relevance, and showing different degrees of climate niche expansion. Four consensus methods were used to combine individual niche models; we assessed transferability of consensus models and the individual models used to generate them. The results suggested that ideal criteria for environmental variable selection vary among algorithms, as different algorithms showed different sensitivities to spatial dimensionality and correlation. Consensus models reflected the central tendency of individual models, and reduced uncertainty by consolidating consistency across individual models, but did not outperform individual models. The question of whether interpolation accuracy comes at the expense of transferability suggests caution in planning methodologies for processing niche models to predict invasive potential. These explorations outline approaches by which to reduce uncertainty and improve niche model transferability with vital implications for ensemble forecasting.

Ecological models used to forecast range change (range change models; RCM) have recently diversified to account for a greater number of ecological and observational processes in pursuit of more accurate and realistic predictions. Theory suggests that process‐explicit RCMs should generate more robust forecasts, particularly under novel environmental conditions. RCMs accounting for processes are generally more complex and data hungry, and so, require extra effort to build. Thus, it is necessary to understand when the effort of building a more realistic model is likely to generate more reliable forecasts. Here, we review the literature to explore whether process‐explicit models have been tested through benchmarking their temporal predictive performance (i.e. their predictive performance when transferred in time) and model transferability (i.e. their ability to keep their predictive performance when transferred to generate predictions into a different time) against simpler models, and highlight the gaps between the rapid development of process‐explicit RCMs and the testing of their potential improvements. We found that, out of five ecological processes (dispersal, demography, physiology, evolution, species interactions) and two observational processes (sampling bias, imperfect detection) that may influence reliability of forecasts, only the effects of dispersal, demography and imperfect detection have been benchmarked using temporally‐independent datasets. Only nine out of twenty‐nine process‐explicit model types have been tested to assess whether accounting for processes improves temporal predictive performance. We found no benchmarks assessing model transferability. We discuss potential reasons for the lack of empirical validation of process‐explicit models. Considering these findings, we propose an expanded research agenda to properly test the performance of process‐explicit RCMs, and highlight some opportunities to fill the gaps by suggesting models to be benchmarked using existing historical datasets.

Fuel Oil Consumption (FOC) represents a significant portion of a fishing vessel’s operating costs, often exceeding 50%. Accurately forecasting FOC during the voyage planning stage is crucial but challenging for optimizing routes and supporting decision-making systems aimed at fuel-saving. Data-driven models have shown excellent performance in FOC prediction. However, gathering the necessary data for these models is expensive and time-consuming. Even though, the applicability of FOC model derived from one vessel to predict FOC for another vessel has received limited research attention. This paper investigates the performance in predicting FOC for an unseen tuna purse seiner, using a two-stage model trained on metocean and operational data, from Copernicus and sensors installed on her similar vessel, respectively. By considering the engine performance modifications, the two-stage model trained on the similar vessel achieves high mean accuracies (over 94%) in predicting FOC for the unseen vessel.