Ignacio Morales-Castilla

Arthropod herbivores cause severe crop yield losses worldwide. Global warming is expected to shift their ranges, phenologies and performances, highlighting the need for improved forecasts of their potential expansion and impacts on food security. After decades of experimental research on thermal biology of arthropod crop pests, available data and tools now allow to model their temperature‐dependent population growth (i.e. population‐level thermal performance curves, or TPCs), which approximate their fundamental thermal niches. Thermal limits of these TPCs can inform biogeographic predictions on latitudinal adaptation to cold and heat, while TPC predictions for population growth rates help assess whether warming will benefit or constrain pest populations. We compiled a dataset on the thermal biology of arthropod crop pests and fitted TPCs to examine the projected shifts in performance under historic and future climates relative to their thermal niches, which were examined through modelling the latitudinal variation of their population‐level thermal limits. Our findings highlight that most arthropod crop pests are projected to increase their thermal potential for population growth under warming by 2050, with temperate populations retaining room to grow but tropical populations increasingly exposed to heat stress. These findings can be explained by the asymmetric variation of thermal niche limits for these intrinsic rates of population increase, since we found that lower thermal limits declined with latitude, while upper thermal limits remained constant.

Background: Phylogenetic imputation—a method that replaces missing trait values with estimates based on evolutionary relationships—is increasingly used for handling data gaps. Although a powerful tool to leverage existing data, it is often treated as a substitute for empirical measurements. Problems and Risks: A growing number of trait datasets report single‐value phylogenetic estimates, often without integrating uncertainty into the primary data products or rigorously evaluating whether the phylogenetic imputation model is informative. For many traits—especially those with weak phylogenetic signal and sparse empirical data—these estimates tend to perform no better than a trivial baseline such as the mean of the observed values. Even when phylogenetic signal is non‐negligible and overall imputation performance is acceptable, improvements at the level of individual estimates are often modest. When used uncritically, such estimates can distort ecological patterns, mislead conservation priorities, and propagate errors across a wide range of downstream applications. Guiding Principles for Imputation: We propose two complementary strategies: (1) prioritising targeted empirical data collection, guided by tools such as phylogenetic ignorance maps to identify critical data gaps, and (2) adopting multiple phylogenetic imputation as a probabilistic framework that quantifies and propagates uncertainty into downstream analyses. We provide practical recommendations for implementing such an uncertainty‐aware workflow, including procedures for assessing how the expected accuracy of phylogenetic estimates varies across individual missing values. Conclusions: Trait‐based science must move beyond the illusion of completeness. Reliable inference depends on clear distinctions between what is known, what is estimated, and how uncertainty influences conclusions. Without appropriate standards and transparency, imputation risks generating misleading results that may ultimately undermine the reliability of subsequent inferences.

Network analysis is a powerful framework for investigating complex systems across disciplines, including ecology. However, ecological network data are often incomplete or error‐prone due to sampling limitations, detection failures and taxonomic uncertainty—leading to missing (false negative) and spurious (false positive) links that obscure structure and hinder inference. The hierarchical stochastic block model (HSBM), particularly in its degree‐corrected form, is among the most effective tools for reconstructing networks under such uncertainty. Despite its robustness, the primary implementation of HSBM in the Python‐based graph‐tool library has remained largely inaccessible to ecologists. Here, we introduce sabinaHSBM , the first R package that makes degree‐corrected HSBM broadly available through a user‐friendly, flexible workflow. By bridging a gap between advanced network modelling and widely used ecological analysis tools, sabinaHSBM facilitates network reconstruction and link prediction from binary data. The workflow involves three main steps: (1) preparing input data, (2) estimating posterior link probabilities and (3) reconstructing the network. The package supports detection of undocumented and spurious links, exploration of hierarchical structure and propagation of uncertainty throughout. Key features include cross‐validation, flexible thresholding, probabilistic evaluation metrics and two link prediction methods: estimating all link probabilities or identifying undocumented ones. We illustrate the package's functionality through a case study using a published global dataset of carnivore–parasite associations, showing that inferred groupings are phylogenetically clustered. To assess predictive accuracy, we examined the top 10 highest probability links identified by the model and found published evidence for eight, despite their absence from the original dataset. This highlights the model's ability to recover biologically meaningful but unrecorded interactions. By providing the first accessible R workflow for HSBM‐based network reconstruction, sabinaHSBM enables researchers to quantify uncertainty and recover overlooked interactions in ecological networks.

Network analysis is a powerful framework for investigating complex systems across disciplines, including ecology, where it helps uncover patterns in predator–prey, host– parasite, or plant–pollinator interactions. However, ecological network data are often incomplete or error-prone due to sampling limitations, detection failures, and taxonomic uncertainty—leading to missing (false negative) and spurious (false positive) links that obscure structure and hinder inference. The hierarchical stochastic block model (HSBM), particularly in its degree-corrected form, is among the most effective tools for reconstructing networks under such uncertainty. Despite its robustness, the primary implementation of HSBM in the Python-based graph-tool library has remained largely inaccessible to ecologists. Here, we introduce sabinaHSBM , the first R package that makes degree-corrected HSBM broadly available through a user-friendly, flexible workflow. By bridging a gap between advanced network modeling and widely used ecological analysis platforms, sabinaHSBM facilitates network reconstruction and link prediction from binary bipartite data. The workflow involves three main steps: (1) preparing input data, (2) estimating posterior link probabilities, and (3) reconstructing the network. The package supports detection of undocumented and spurious links, exploration of hierarchical structure, and propagation of uncertainty throughout. Key features include cross-validation, flexible thresholding, probabilistic evaluation metrics, and two link prediction modes: estimating all link probabilities or identifying undocumented ones. We illustrate the package’s functionality through a case study using a published global dataset of carnivore–parasite associations, showing that inferred groupings are phylogenetically clustered. To assess predictive accuracy, we examined the top 10 highest-probability links identified by the model and found published evidence for 8, despite their absence from the original dataset. This highlights the model’s ability to recover biologically meaningful but underreported interactions. By integrating all components of HSBM-based reconstruction into an accessible R package, sabinaHSBM empowers researchers to improve relational data quality and uncover overlooked patterns in complex ecological networks and beyond.

Los espacios protegidos de montaña son áreas en las que el turismo es un importante recurso económico, su gestión integra prácticas ganaderas extensivas que coexisten con la fauna salvaje. Esta relación entre ganado, fauna salvaje y personas puede provocar problemas sanitarios relacionados con zoonosis pudiendo afectar a la salud pública, la biodiversidad y la economía. Múltiples zoonosis son transmitidas por ectoparásitos como garrapatas que actúan como reservorio y vector de enfermedades infecciosas siendo los principales hospedadores pequeños mamíferos, ungulados silvestres y ganado. El propósito de este proyecto es plantear la creación de modelos de abundancia de vectores y patógenos teniendo en cuenta la importancia del agua como determinante del uso del espacio. Se están realizando muestreos en el Parque Nacional de Ordesa y Monte Perdido, el Parque Nacional de la Sierra de Guadarrama y el Parque Natural de las Sierras de Cazorla Segura y las Villas; midiendo: la densidad de garrapatas/m2 en diferentes hábitats y distancias al agua; la densidad de ungulados salvajes y domésticos utilizando cámaras trampa, censos visuales en el crepúsculo y entrevistas a ganaderos y personal del parque; la densidad de pequeños mamíferos mediante métodos de captura y recaptura. Las muestras recogidas se analizarán en laboratorio preparando bibliotecas genómicas y análisis matriciales, que permitirán construir redes probabilísticas a partir de las cuales realizar mapas de riesgos que podrían servir de base para la gestión, mitigando y previniendo la transmisión de enfermedades, protegiendo la economía y la salud de la población.

Anthropogenic climate change has uneven impacts across the globe and throughout the year. Such unevenness poses a major challenge for human adaptation, especially for agricultural and other managed systems. Estimating effects beyond one region is challenging, however, because differences between regions make it difficult to know what seasonal periods of climate to compare. Both local climate and the planting decisions of a region affect the relevant periods for estimating changes in climate. Here, we use recent phenological models with a dataset of mean phenology for over 500 cultivars (varieties) to estimate climatic changes in growing regions across the globe for a major perennial crop that has been highly affected by climate change: winegrapes. We examine a suite of grower-relevant metrics, including temperatures during budburst, throughout the growing season and temperatures and precipitation surrounding harvest. We find that climate change has impacted all regions, especially for heat metrics across the full growing season (GDD, maximum temperature and days above 35°C). By far the largest shifts, however, are in European regions, where the number of hot days (>35°C) and maximum growing season temperatures are several standard deviations higher than before significant anthropogenic climate change. Including variety diversity in our estimates impacted only metrics at the start and end of the season, appearing most important for harvest-related climate metrics, and then only in ‘Old World’ regions, where most variety diversity is planted. Climate change impacts have thus been highly uneven across the world’s winegrowing regions and the impacts are variable across the growing season. Navigating how best to adapt the global winegrowing industry to climate change will require addressing these spatial and temporal complexities.

Understanding how temperature determines the distribution of life is necessary to assess species’ sensitivities to contemporary climate change. Here, we test the importance of temperature in limiting the geographic ranges of ectotherms by comparing the temperatures and areas that species occupy to the temperatures and areas species could potentially occupy on the basis of their physiological thermal tolerances. We find that marine species across all latitudes and terrestrial species from the tropics occupy temperatures that closely match their thermal tolerances. However, terrestrial species from temperate and polar latitudes are absent from warm, thermally tolerable areas that they could potentially occupy beyond their equatorward range limits, indicating that extreme temperature is often not the factor limiting their distributions at lower latitudes. This matches predictions from the hypothesis that adaptation to cold environments that facilitates survival in temperate and polar regions is associated with a performance trade-off that reduces species’ abilities to contend in the tropics, possibly due to biotic exclusion. Our findings predict more direct responses to climate warming of marine ranges and cool range edges of terrestrial species.

Understanding how temperature determines the distribution of life is necessary to assess species’ sensitivities to contemporary climate change. Here we test the importance of temperature in limiting geographic ranges of ectotherms by comparing temperatures across occupied ranges to those species could potentially occupy based on their physiological thermal tolerances. Whereas marine and tropical terrestrial species occupy temperatures that closely match their thermal tolerances, high-latitude terrestrial species under-occupy warm temperatures and are absent from thermally tolerable areas towards the equator. This suggests that on land, temperature less often limits the equatorward range edge of temperate species, supporting the hypothesis that their historic expansion ‘out of the tropics’ was associated with tropical biotic exclusion. Our findings predict more direct responses to climate warming of marine ranges and cool range edges of terrestrial species.