Context: Recent large and high-severity wildfires have burned vast areas of coniferous forests throughout Western North America. These burned landscapes are recovering amid increasingly frequent climate extremes, such as drought. We need to understand how post-fire climate extremes and other ecological drivers (such as fire impacts) influence patterns and trends of coniferous recovery.
Objectives: We worked at a landscape scale (> 400,000 hectares) to investigate the association between distinct post-fire forest recovery and ecological drivers in dry sub-boreal forests. We created structural recovery groups distinct in patterns and trends of coniferous cover and density and then modeled their association with ecological drivers.
Methods: We used Landsat time-series data to identify unique spectral recovery, which we grouped based on post-fire regrowth and stocking estimates. Remotely Piloted Aircraft light detection and ranging (lidar) provided structural estimates 5-21 years post-fire. We modeled the association between structural recovery groups and ecological drivers with random forests. For each category of drivers (site conditions, climate, climate anomalies, pre-fire composition, and fire impacts), we used individual models to identify important drivers. We then incorporated the most important drivers in a global model to highlight the drivers that were important across categories.
Results: Initial spectral trends indicated longer-term differences in structural forest recovery. Climate anomalies (such as post-fire extremes in temperature and precipitation) and pre-fire basal area best predicted observed structural groupings-abnormally cold and dry summers after the fire were associated with slow conifer establishment. Comparatively, areas with a higher pre-fire basal area maintained a mixed canopy of deciduous and coniferous stems.
Conclusions: At a landscape scale, post-fire climate conditions best predicted structural forest recovery, suggesting management plans should be adaptable to the conditions experienced post fire.
Supplementary information: The online version contains supplementary material available at 10.1007/s10980-025-02266-y.
Context: While biodiversity decline is undebated on the global level, landscape-scale trends are poorly known and local assemblages even show stable species richness, accompanied by pronounced turn-over. The landscape-scale consequences of local-scale species turnover likely depend on whether species replacement is random or biased towards more frequent species in the metacommunity, but this potential bias is insufficiently studied.
Objectives: Here, we use grassland ecosystems of a Central European national park to simultaneously analyse time-series of local-scale species richness and landscape-scale species incidence to better understand how trends are linked at these two scales.
Methods: From 2013 to 2024 we sampled 120 plots per year and used regression methods to quantify changes in the number of species per assemblage, the incidence of species across assemblages and the relationship between initial incidence of species and incidence trends. To explore possible drivers of change, we further evaluated trends of community means of environmental indicator values.
Results: We found that local species richness has increased within the study period from 18 species per plot in 2013 to 21 species in 2024, while the overall number of species sampled per year stayed the same. In contrast, when looking at individual species trends we found an average decline of species' incidence in the region. While a small pool of already common species became more frequent, the majority of species became rarer, leading to a pronounced homogenization of plant communities on the sampled sites. Indicator-value analysis showed that the species turnover was mainly influenced by desiccation of grasslands, significantly biassing incidence changes towards species that prefer drier conditions.
Conclusions: We conclude that in typical Central-European grassland ecosystems, anthropogenic drivers rather decrease landscape-scale than local-scale biodiversity, because they tend to homogenize environmental conditions. The resulting species turn-over can stabilize local species richness but depletes the metacommunity, thereby posing future risks to the regional biodiversity.
Supplementary information: The online version contains supplementary material available at 10.1007/s10980-025-02256-0.
Context: Understanding the roles of different drivers in land use and land cover change (LULCC) is a critical research challenge. However, as LULCC is the result of complex, socio-ecological processes and is highly context dependent, achieving such understanding is difficult. This is particularly true for causal modelling approaches that are critical for effective policy formulation. Causal machine learning (ML) methods could help address this challenge, but are as yet poorly understood or applied by the LULCC community.
Objectives: To provide an accessible introduction to the state of the art for causal ML methods, their limitations, and their potential applications understanding LULCC.
Methods: We conducted two workshops where we identified the most promising ML methods for increasing understanding of LULCC dynamics.
Results: We provide a brief overview of the challenges to causal modelling of LULCC, including a simple example, and the most relevant causal ML approaches for addressing these challenges, as well as their limitations.
Conclusions: Causal ML methods hold considerable promise for improving causal modelling of LULCC. However, the complexity of LULCC dynamics mean that such methods must be combined with domain understanding and qualitative insights for effective policy design.
Context: Theoretical research has considered how animals should optimise foraging strategies to maximise fitness, adapting search scale to exploit different habitats and minimise competition. Empirical studies have described multi-scale area-restricted search (ARS) strategies for some species, but the physical and biological mechanisms underpinning such behaviour are rarely studied.
Objectives: Our objectives were to quantify the presence, prevalence, and habitat associations of scale-dependent foraging for two sympatric seal species, accounting for regional variation across the seascape.
Methods: We analyse a GPS telemetry dataset of 116 grey (Halichoerus grypus) and 325 harbour seals (Phoca vitulina) tracked throughout the North Sea. We test the existence of multi-scale ARS, comparing hidden Markov models (HMMs) with two ARS states against more conventional HMMs (one ARS state). We quantify regional variation and examine the scale-dependence of foraging habitat associations using post-hoc "use-encounter" models.
Results: Both species exhibited nested broad-scale and focussed ARS. Accounting for scale resulted in increases of up to 25% and 46% in inferred ARS for grey and harbour seals respectively. The prevalence and habitat associations of different ARS scales varied in a regional species-specific manner.
Conclusions: We demonstrate the first application of HMMs to capture multi-scale ARS from animal-borne tracking data. Overlooking scale-dependence may mask individual variation and underestimate ARS, with consequences for ecological understanding and conservation applications. We hypothesise that seals employ different search scales for different habitats, competition levels and/or prey types. We call for further research to elucidate the prevalence and ecological significance of this phenomenon in other aquatic predators.
Supplementary information: The online version contains supplementary material available at 10.1007/s10980-025-02281-z.
Context: Connectivity across river networks facilitates species movement and ecological processes that contribute to freshwater biodiversity. Certain indices provide measures of connectivity to focus conservation planning.
Objectives: Our objective was to test whether commonly used connectivity indicators based on network structure can reliably predict population persistence.
Methods: We used a spatially explicit metapopulation model for freshwater fish that complete their life cycle entirely within river networks and depend on connectivity for movement. Simulations were conducted across a range of network sizes, topologies, dispersal abilities, and barrier passabilities. We assessed the relationship between the Dendritic Connectivity Index (DCI) and metrics of persistence at the network and the reach scale.
Results: DCI was strongly correlated with persistence at both the network and reach scale across most simulated network sizes and configurations, particularly in dendritic (branching) systems with symmetric barrier passability. At the network scale, correlations were strongest with density-independent persistence metrics, which is expected since DCI does not incorporate population interactions. Species dispersal ability influenced DCI-persistence correlations differently across scales: correlations were strongest at the network scale when dispersal distances spanned the full network (global dispersal) and at the reach scale when movement was limited to neighbouring segments (local dispersal). We also found that increases in DCI following simulated barrier removal were associated with improvements in persistence, further demonstrating its potential to support restoration efforts.
Conclusion: Indicators like DCI can inform connectivity-focused conservation planning in river networks.
Supplementary information: The online version contains supplementary material available at 10.1007/s10980-025-02278-8.
Context: The > 25,000 km2 Flint Hills ecoregion in eastern Kansas and northeastern Oklahoma, USA, is an economically and ecologically important area encompassing the largest remaining tallgrass prairie ecosystem in North America. Prescribed fires are used routinely to control invasive woody species and improve forage production for the beef-cattle industry. However, burning releases harmful pollutants that, at times, contribute to air quality problems for communities across a multi-state area.
Objectives: Establish a modeling framework for synthesizing long-term ecological data in support of Flint Hills tallgrass prairie management goals for identifying how much, where, and when rangeland burning can be conducted to maximize ecological and economic benefits while minimizing regional air quality impacts.
Methods: We used EPA's VELMA ecohydrology model to synthesize long-term experimental data at the 35 km2 Konza Prairie Biological Station (KPBS) describing the effects of climate, fire, grazing, topography, and soil moisture and nutrient dynamics on tallgrass prairie productivity and fuel loads; and to spatially extrapolate that synthesis to estimate grassland productivity and fuel loads across the nearly 1000 times larger Flint Hills ecoregion to support prescribed burning smoke trajectory modeling using the State of Kansas implementation of the U.S. Forest Service BlueSky framework.
Results: VELMA provided a performance-tested synthesis of KPBS data from field observations and experiments, thereby establishing a tool for regionally simulating the combined effects of climate, fire, grazing, topography, soil moisture, and nutrients on tallgrass prairie productivity and fuel loads. VELMA's extrapolation of that synthesis allowed difficult-to-quantify fuel loads to be mapped across the Flint Hills to support environmental decision making, such as forecasting when, where, and how prescribed burning will have the least impact on downwind population centers.
Conclusions: Our regional spatial and temporal extrapolation of VELMA's KPBS data synthesis posits that the effects of integrated ecohydrological processes operate similarly across tallgrass prairie spatial scales. Based on multi-scale performance tests of the VELMA-BlueSky toolset, our multi-institution team is confident that it can assist stakeholders and decision makers in realistically exploring tallgrass prairie management options for balancing air quality, tallgrass prairie sustainability, and associated economic benefits for the Flint Hills ecoregion and downwind communities.
Context: Land-use intensification to increase yields is often detrimental to biodiversity undermining the provision of ecosystem services. However, it is questionable if ecosystem service providers contribute to ecological intensification by achieving the same or higher yields than conventional high-intensity agriculture.
Objectives: In this study, we aimed to disentangle the effects of local and landscape-scale land-use intensification on arthropod communities and their contribution to ecosystem services and crop yield. A set of meta-analytic structural equation models allowed us to assess direct and indirect relationships in the cascade from land use to yield.
Methods: We selected 37 datasets containing information on land use, community composition, levels of pollination and natural pest control services, and crop yield. We quantified functional diversity of communities by collecting trait information for three exemplary groups of service-providers: bees, ground beetles, and spiders.
Results: Local land-use intensification reduced the abundance of all arthropod groups. Spiders were the only group whose species richness was negatively related to a higher percentage of arable land in the landscape. High abundance of bees related positively to oilseed rape pollination and crop yields. In the models for the two predator groups, crop yield was strongly determined by land use, independent of the pest control services provided by natural enemies.
Conclusions: Our results suggest a potential for ecological intensification mediated by land-use change in crops where pollination benefits yield, but suggest more nuanced effects for pest control. Our study also calls for experiments on multiple taxonomic groups and ecosystem services that apply comparable methods at similar scales.
Supplementary information: The online version contains supplementary material available at 10.1007/s10980-025-02117-w.
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：