Ecological Research最新文献

Food web robustness is a critical aspect of ecosystem stability and has been extensively studied in ecology. However, the potential of machine learning techniques in predicting food web robustness and the identification of key network structure indicators have not been fully explored. We compared the suitability of different machine learning methods and assessed the relative importance of network structure indicators for predicting the robustness of food webs. We utilized a variety of food web datasets spanning different ecosystems to calculate network structure indicators, which include average distance (AD), betweenness centrality (BC), directional connectivity (C), closeness centrality (CC), diameter (D), degree centrality (DC), edge betweenness centrality (EBC), number of links (L), linkage density (LD), and number of nodes (N). We then compared the performance of machine learning methods, including artificial neural network (ANN), random forest (RF), least absolute shrinkage and selection operator (LASSO), and decision tree (DT), and evaluated the relative importance of network structure indicators on robustness predictions. The results demonstrate that the RF model has the best performance (MAE = 0.0178, RMSE = 0.0263, R² = 0.9063). Meanwhile, the CC indicator has a significant impact in predicting robustness of food webs. It is suggested that both the RF model and the CC indicator should be considered seriously in predicting food web robustness. This research elucidates the differential outcomes when various machine learning methodologies and indicators are employed to predict the robustness of food webs. It significantly enhances our understanding by demonstrating the precise capability of machine learning models in forecasting the robustness of food webs.

Vascular plants develop diverse root system architectures and exudates to facilitate acquisition of soil nutrients (nitrogen and phosphorus). Plant species have inherently specific properties of root system architectures and exudates, but some plants exhibit high plasticity to respond to spatiotemporal variations in soil nutrient forms and availability. This paper synthesizes evidence on how plant species diversity and root plasticity contribute to soil nutrient mobilization and uptake in plants from the tropics to the Arctic with varying plant species diversity. The development of finer roots in the surface soil is a well-known strategy for the acquisition of limited nutrients (especially phosphorus), but the allocation of roots foraging “nutrient hotspots” in deeper soil (podzolic soil or permafrost-affected soils) is an alternative strategy for coniferous trees tested in the Arctic and boreal forests. This contrasts with findings in tropical forests, where diverse plant species with different capacities of organic acid exudation coexist and species composition shifts in response to phosphorus deficiency. In particular, high malate exudation from roots and rhizosphere microbes stimulates phosphorus solubilization, aluminum detoxification, and lignin degradation in acidic soils. The diversity and plasticity of the root system architecture, root exudation, and the flexibility of nutrient sources mitigate nutrient limitation in soil. Root plasticity facilitating soil nutrient acquisition has a large impact on biogeochemistry and soil formation, such as podzolization, in the long term.

Hyperaccumulator plants are a botanical curiosity that have allowed the development of agromining of metals, with a special focus on nickel. In nickel agromining, this element is recovered from ashed hyperaccumulators cultivated on metal-rich soils. In order to explore bio-based approaches for the decomposition of hyperaccumulator biomass and nickel recovery that do not include burning, we performed a vermicomposting experiment using the earthworm species Eisenia andrei and the biomass of Bornmuellera emarginata (which contained almost 1% of nickel). We conducted our experiment for 12 weeks and assessed the decomposition process of the hyperaccumulator biomass, changes in earthworm number and biomass, and changes in nickel concentration and mobility. Despite the initial mortality and an increase of Ni concentration in earthworm tissues, E. andrei was able to decompose B. emarginata biomass. This process also showed a massive colonization of the biomass by a fungus during the first weeks of the assay. Our results indicate that the vermicomposted hyperaccumulator biomass had a higher nickel concentration than the starting material but the diethylenetriaminepentaacetic acid-extractable nickel decreased. At the same time, due to earthworm activity, the nickel was redistributed and diluted in the vermicompost bedding, reducing the interest of this approach for agromining, but opening the perspective of using the vermicomposted hyperaccumulator biomass as an organic amendment in nickel-deficient crops.

Although checklists are essential tools for managing and conserving ultramafic ecosystems, no updated checklist currently exists for ultramafic ecosystems on the southern Iberian Peninsula. Thus, the objectives of our study were (1) to create an updated checklist of serpentinophytes on southern Iberian Peninsula, (2) to determine whether the distribution of serpentinophytes is associated with certain specific types of ultramafic rocks, and (3) to calculate the abundance and richness of serpentinophytes per outcrop to guide conservation efforts. Following a methodology involving field work and searches of bibliographies and herbaria we produced an updated checklist containing 28 serpentinophytes (i.e., 23 obligates, one preferential taxon, one sub-serpentinophyte, and three regional serpentinophytes). The serpentinophytes showed different petrological affinity, where harzburgite–lherzolite and harzburgite–pyroxenite–dunite exhibited higher occupancy, possibly due to their mineralogical and chemical composition (e.g., containing heavy metals) and/or the larger surface area of those outcrops. We also observed that the occupancy of 21 species was higher in different petrographic entities, the reasons for which could be elucidated by future soil analyses. The highest richness of serpentinophytes was found in the main outcrop of Bermeja, followed by smaller outcrops of Alpujata, Aguas, and Guadalhorce Valley. Although the richness of Aguas resembled that of Alpujata, a notable difference emerged in some of its areas owing to bioclimatic and biogeographic isolation. Given the exclusive presence of serpentinophyte flora on the southern Iberian Peninsula, all southern Iberian outcrops should receive some form of conservation as protected natural spaces.

Noccaea praecox and Noccaea caerulescens (Brassicaceae) are nickel and zinc hyperaccumulators, native to Europe. To date, most studies have focused on metal accumulation in the leaves, whereas the distribution of metals in the inflorescences of hyperaccumulator plants remains largely unexplored, but of great interest in the context of adaptation to fertility and (insect) pollination. Samples of N. praecox from an ultramafic site and N. caerulescens from an industrially contaminated site were used for synchrotron-based micro-X-ray fluorescence (μXRF) analysis. The results showed that nickel and zinc in the flowers of N. praecox are mainly distributed in the receptacle, ovary, and anthers, but at different concentrations. Similar results were found in N. caerulescens, with the greatest accumulation in the receptacle and ovary, especially in the walls, however at lower levels in the anthers. Although the leaves of N. praecox and N. caerulescens are the main deposition sites for nickel and zinc, significant concentrations of these elements were deposited in the flowers, especially in the pistils and anthers, indicating possible negative effects on fertility and pollinator species.

One of the challenges of agromining is the adoption of more environmentally-friendly solutions to improve plant biomass yields and Ni concentrations in plants. Here, we focused on four sustainable solutions for optimizing nickel phytoextraction by the hyperaccumulator Odontarrhena chalcidica: a biostimulant, another biostimulant/plant defense stimulator, artificial root exudates, and a biodegradable metal chelator. Their effects on the growth and physiology of O. chalcidica, on Ni phytoextraction capacity, on physicochemical soil characteristics, and on the diversity of rhizosphere and endophytic bacteria were compared to a conventional mineral fertilizer. A 5-month pot experiment was carried out with O. chalcidica growing on an ultramafic soil. Element concentrations in both soil and plant were measured. Moreover, numerous compounds were analyzed (photosynthetic pigments, malondialdehyde, flavonoids, free amino acids, and starch). We also characterized rhizosphere and endophytic bacterial communities associated with this hyperaccumulator. Biostimulants appeared to be a promising way of improving Ni concentration in shoots and plant biomass production, and showed a positive effect on bacterial richness and diversity. In contrast, our experiments did not show that artificial exudates and mineral fertilizer had a positive effect on Ni phytoextraction. Finally, the biodegradable chelator had no significant effect. The use of sustainable amendments into a Ni agromining system improved both plant biomass and Ni yields, in comparison to mineral fertilization. Thus, improving agromining by replacing mineral fertilizers would be an eco-efficient strategy.

Land cover classification mapping is the process of assigning labels to different types of land surfaces based on overhead imagery. However, acquiring reference samples through fieldwork for ground truth can be costly and time-intensive. Additionally, annotating high-resolution satellite images poses challenges, as certain land cover types are difficult to discern solely from nadir images. To address these challenges, this study examined the feasibility of using street-level imagery to support the collection of reference samples and identify land cover. We utilized 18,022 images captured in Japan, with 14 different land cover classes. Our approach involved using convolutional neural networks based on Inception-v4 and DenseNet, as well as Transformer-based Vision and Swin Transformers, both with and without pre-trained weights and fine-tuning techniques. Additionally, we explored explainability through Gradient-Weighted Class Activation Mapping (Grad-CAM). Our results indicate that using a Vision Transformer was the most effective method, achieving an overall accuracy of 86.12% and allowing for full explainability of land cover targets within an image. This paper proposes a promising solution for land cover classification from street-level imagery, which can be used for semi-automatic reference sample collection from geo-tagged street-level photos.