Explore the vast range of titles available.

Accurate segmentation of fine roots in field rhizotron imagery is essential for high-throughput root system analysis but remains challenging due to limitations of traditional methods. Traditional methods for root quantification (e.g., soil coring, manual counting) are labor-intensive, subjective, and low-throughput. These limitations are exacerbated in in situ rhizotron imaging, where variable field conditions introduce noise and complex soil backgrounds. To address these challenges, this study develops an advanced deep learning framework for automated segmentation. We propose an improved U-shaped Convolutional Neural Network (U-Net) architecture optimized for segmenting larch (Larix olgensis) fine roots under heterogeneous field conditions, integrating both in situ rhizotron imagery and open-source multi-species minirhizotron datasets. Our approach integrates (1) a Convolutional Block Attention Module (CBAM) to enhance feature representation for fine-root detection; (2) an additive feature fusion strategy (UpAdd) during decoding to preserve morphological details, particularly in low-contrast regions; and (3) a transfer learning protocol to enable robust cross-species generalization. Our model achieves state-of-the-art performance with a mean intersection over union (mIoU) of 70.18%, mean Recall of 86.72%, and mean Precision of 75.89%—significantly outperforming PSPNet, SegNet, and DeepLabV3+ by 13.61%, 13.96%, and 13.27% in mIoU, respectively. Transfer learning further elevates root-specific metrics, yielding absolute gains of +0.47% IoU, +0.59% Precision, and +0.35% F1-score. The improved U-Net segmentation demonstrated strong agreement with the manual method for quantifying fine-root length, particularly for third-order roots, though optimization of lower-order root identification is required to enhance overall accuracy. This work provides a scalable approach for advancing automated root phenotyping and belowground ecological research.

Increases in carbon (C) inputs to soil can replenish soil organic C (SOC) through various mechanisms. However, recent studies have suggested that the increased C input can also stimulate the decomposition of old SOC via priming. Whether the loss of old SOC by priming can override C replenishment has not been rigorously examined. Here we show, through data–model synthesis, that the magnitude of replenishment is greater than that of priming, resulting in a net increase in SOC by a mean of 32% of the added new C. The magnitude of the net increase in SOC is positively correlated with the nitrogen-to-C ratio of the added substrates. Additionally, model evaluation indicates that a two-pool interactive model is a parsimonious model to represent the SOC decomposition with priming and replenishment. Our findings suggest that increasing C input to soils likely promote SOC accumulation despite the enhanced decomposition of old C via priming. The magnitudes of replenishment and priming, two important but opposing fluxes in soil organic carbon (SOC) dynamics, have not been compared. Here the authors show that the magnitude of replenishment is greater than that of priming, resulting in a net SOC accumulation after additional carbon input to soils.

Root turnover is a key process contributing to soil carbon storage, nutrient cycling, and other ecosystem functions. However, quantifying root turnover rates remains highly uncertain and methodologically challenging. Field rhizotrons were employed to quantify root turnover times using median longevities of five branching orders in a Larix gmelinii plantation. Root images were recorded by scanning the rhizotron windows at a monthly interval during four growing seasons. Root demographic data and branching orders were obtained by analyzing these images using Rootfly software coupled with manual mouse‐tracing of individual roots. Root longevities and turnover estimates were calculated using these data. Roots of different branching orders showed significantly different turnover times. The mean turnover times of the first‐order roots and second‐order roots were 284 and 994 d, respectively. Roots of higher branching orders (third to fifth orders) remained alive at the end of the 4‐yr experimental period, indicating much longer turnover times than the duration of the experiment. Root turnover times increased exponentially with branching orders. Further analysis of these data suggested that root branching orders combined with sampling biases, timing of root cohorts, and longevity distribution patterns crucially influenced root turnover times. The method of combining field glass rhizotrons with electronic scanning permits quantification of root turnover for five branching orders in a temperate forest. The overall result empirically demonstrates the crucial role of branching orders for accurately quantifying root turnover times.

Aims Rhizosphere is a hotspot for soil C and N biogeochemical cycling in terrestrial ecosystems. However, the interaction between soil C and N mineralization remain poorly understood in the rhizosphere soils. This study aimed to identify interactions between soil C and N mineralization in rhizosphere soils and bulk soils at a large scale. Methods We used the “soil adhering to fine roots after shaking” method to collect paired rhizosphere soils and bulk soils along an altitudinal forest gradient. Soil C mineralization rates (C min ) and net N mineralization rates (N min ) were determined with laboratory incubation. Results We found a significantly positive relationship between C min and N min in the rhizosphere soils across sites, whereas C min was not correlated with N min in the bulk soils. Furthermore, soil properties, microbial biomass C (MBC) and extracellular enzyme activities showed substantial paths affecting C min and N min using structural equation model. The coupling of C min and N min in rhizosphere soils could be triggered by root-soil interactions, resulting in the higher level of MBC, total organic C of soil, total N of soil, and extracellular enzyme activities. By contrast, the decoupling of C min and N min in the bulk soils might be attributed to the lower level of MBC and extracellular enzyme activities. Conclusions Our results demonstrated that soil C and N mineralization coupled in the rhizosphere rather than in the bulk soils. These results suggest that the interaction between soil C and N cycling in rhizosphere are likely to differ from that in bulk soil.

Fine roots play a key role in carbon, nutrient, and water biogeochemical cycles in forest ecosystems. However, inter-annual dynamics of fine root production, mortality, and turnover on the basis of long-term measurement have been less studied. Here, field scanning rhizotrons were employed for tracking fine root by branch order over a 6 years period in a larch plantation. For total fine roots, from the first- to the fifth-order roots, annual root length production, length mortality, standing crops, and turnover rate varied up to 3.4, 2.3, 1.5, and 2.3-folds during the study period, respectively. The inter-annual variability of those roots indices in the first-order and the second-order roots were greater than that of the higher order (third- to fifth-order) roots. The turnover rate was markedly larger for the first-order roots than for the higher order roots, showing the greatest variability up to 20 times. Seasonal dynamics of root length production followed a general concentrated pattern with peak typically occurring in June or July, whereas root length mortality followed a general bimodal mortality pattern with the dominant peak in May and the secondary peak in August or October. Furthermore, the seasonal patterns of root length production and mortality were similar across years, especially for the first-order and the second-order roots. These results from long-term observation were beneficial for reducing uncertainty of characterizing fine root demography in consideration of large variation among years. Our findings highlight it is important for better understanding of fine root dynamics and determining root demography through distinguishing observation years and root branch orders.

Mixed plantation of Chinese fir (Cunninghamia lanceolata) is an effective artificial forest management for tree productivity. However, the mixing strategies, site conditions, and subsurface properties that affect tree productivity are not yet fully understood. In this study, we conducted a meta-analysis of 96 publications to consolidate insights on the effects of mixing strategies (e.g., planting density, mixing proportion, mixed species, and tree age), site conditions (e.g., mean annual precipitation (MAP), mean annual temperature (MAT), elevation, and total nitrogen (TN) or total phosphorus (TP) of sample sites), and subsurface properties (e.g., soil characteristics, microbial communities, and extracellular enzyme activity) on tree height, diameter at breast height, and individual volume of Chinese fir. We used the Web of Science and China National Knowledge Infrastructure for searching peer-reviewed papers, and the searching words were: (“Cunninghamia lanceolata” OR “Chinese fir”) AND “mix*”. Following the data screening process, the natural logarithm of the response ratio (lnRR) was computed for subsequent analysis. The results showed that introduced companion species generally increased the individual volume of Chinese fir by an average of 20%. Densities ranging from 1200 to 2000 trees per hectare and moderate mixing proportions (1:1 to 3:1) optimized individual tree growth and thereby boosted productivity. Broadleaf species may be beneficial companions, and trees aged 10 to 20 years grew fastest. At sites with low MAT and high MAP, mixed plantations enhanced the tree productivity of Chinese fir. The optimal elevation range for mixed plantations may be 200 to 600 m. Further, mixed plantations significantly changed soil properties by improving soil structure, increasing soil pH and soil water content, and soil total and available N and P, which were crucial for boosting the productivity of Chinese fir. Soil microbial biomass and enzyme activities were also significantly increased by mixed plantations. Overall, these findings highlight the importance of mixing strategies and site conditions in increasing tree productivity of Chinese fir by improving soil physicochemical characteristics, increasing resource availability, and reducing interspecific and intraspecific competition through niche separation.

Forest gap models are important tools for assessing the impact of global climate change on forest dynamics of tree species composition and size structure. In this study, the FAREAST gap model was used to examine the response of forest dynamics on Gongga Mountain, which is located on the southeastern fringe of the Tibetan Plateau, under three climate change scenarios. The simulated results showed that the climax community of the deglaciation slash would be mixed species of Picea brachytyla, Tsuga chinensis, and Pinus densata under climate change scenarios, as opposed to the pure Abies fabri forest under the current climate. Climate change also drove replacement of Populus purdomiis by Betula utilis, which became the most abundant pioneer tree species on the deglaciation slash. Under scenarios of climate change, three responses of the four typical forests distributed between 2200 and 3580 m above sea level are observed, such as dieback of today's forest at 2200 and 3150 m, gradual change of the species composition at 2780 m, and afforestation at 3580 m. It is worth noting that the scenarios of climatic change are of inherent uncertainty, in the same way as the formulation of the ecological factors used in the models. It is suggested that simulations not be interpreted as predictions of the future development of the forest, but as a means of assessing their sensitivity to climate change. It is concluded that mountainous forests are quite sensitive to climate change.

Rhizosphere effects (REs) play important roles in regulating carbon (C) and nutrient cycling in terrestrial ecosystems. However, little is known about the REs of mature trees in the field, especially at the ecosystem scale. This study aimed to explore the variation and patterns of REs in natural ecosystems. Here, combining soil monoliths with an adhering soil (shaking fine roots) method was adopted to sample paired rhizosphere soil and bulk soil along an altitudinal gradient. Based on the relative REs and the percentage of rhizosphere soil mass, the REs on soil C and net nitrogen mineralization rates (Cmin and net Nmin) at the ecosystem scale were estimated. Our results showed that the REs on soil processes, soil microbial biomass C and extracellular enzyme activities (β-glucosidase and N-acetyl-glucosaminidase activities), and soil chemical properties (total C, total N, inorganic N, extractable P, K, Ca, Mg, Fe, and Mn) were significantly positive across altitudinal sites, while soil pH was significantly negative. Although the relative REs on investigated variables varied significantly among altitudes, the relative REs did not show a clear trend with the increased altitudes. Across altitudes, the mean magnitude of ecosystem-level REs on Cmin and net Nmin were 19% (ranging from 4% to 48%) and 16% (ranging from 3% to 34%), respectively. Furthermore, the magnitude of ecosystem-level rhizosphere effects increased linearly with the increased altitudes. The altitudinal patterns of ecosystem-level RE mainly depend on the percentage of rhizosphere soil mass. In conclusion, our results provided a set of new evidence for the REs, and highlighted the need to incorporate REs into land C and N models.

Layered materials exhibit different electronic and phonon properties along in-plane and out-of-plane directions; existing studies focus on their in-plane behaviors, and the influence of such anisotropies on the dynamics of photocarriers and phonons is unknown. Here, we fabricate layered PdSe 2 crystals with flat edge surfaces and compare the time-resolved ultrafast spectroscopies on their basal and edge surfaces. Pronounced differences in the transient reflection spectroscopies reveal the inconsistent photocarrier and phonon dynamics behaviors on the two surfaces: the slow hot carrier relaxation process is accelerated and the thermoelasticity-induced longitudinal coherent acoustic phonon oscillation completely vanishes on the edge surface, as compared with the basal surface. Theoretical analysis reveals that the inconsistent hot carrier dynamics originate from the anisotropic properties of low-energy phonons in PdSe 2 , and the absence of phonon oscillation on the edge surface results from the wavevector-limited sensitivity of acoustic B 1 u mode. Moreover, polarization-dependent spectroscopies indicate the diverse optical anisotropies beyond the in-plane of PdSe 2 . This work provides a new method to explore unique physical properties and modulate the optical anisotropy of layered materials.

The process-oriented model Forest-DNDC describing biogeochemical cycling of C and N and GHGs (greenhouse gases) fluxes (CO 2 , NO and N 2 O) in forest ecosystems was applied to simulate carbon sequestration and GHGs emissions in Abies fabric forest of the Gongga Mountains at southeastern edge of the Tibetan Plateau. The results indicated that the simulated gross primary production (GPP) of Abies fabric forest was strongly affected by temperature. The annual total GPP was 24,245.3 kg C ha −1 yr −1 for 2005 and 26,318.8 kg C ha −1 yr −1 for 2006, respectively. The annual total net primary production (NPP) was 5,935.5 and 4,882.2 kg C ha −1 yr −1 for 2005 and 2006, and the annual total net ecosystem production (NEP) was 4,815.4 and 3,512.8 kg C ha −1 yr −1 for 2005 and 2006, respectively. The simulated seasonal variation in CO 2 emissions generally followed the seasonal variations in temperature and precipitation. The annual total CO 2 emissions were 3,109.0 and 4,821.0 kg C ha −1 yr −1 for 2005 and 2006, the simulated annual total N 2 O emissions from forest soil were 1.47 and 1.36 kg N ha −1 yr −1 for 2005 and 2006, and the annual total NO emissions were 0.09 and 0.12 kg N ha −1 yr −1 for 2005 and 2006, respectively.