Forest Systems Research Articles

Deciphering the vectors: Unveiling the local dispersal of Litylenchus crenatae ssp. mccannii in the American beech (Fagus grandifolia) forest ecosystem

Beech leaf disease (BLD), caused by the Litylenchus crenatae ssp. mccannii (Lcm) nematode, is an emerging threat to beech trees. This disease is characterized by distinct leaf symptoms, including leaf interveinal banding and thickened leaf texture, which leads to eventual tree mortality. Understanding Lcm dispersal mechanism(s) is crucial for managing BLD, yet these remain largely unknown, posing a major barrier to its effective management. This study represents a pioneering investigation into the abiotic and biotic vectors that potentially contribute to the local dispersal of Lcm in natural American beech (Fagus grandifolia) forest systems in the Northeastern United States. An experiment was set up in Stone Valley Forest, Pennsylvania (PA), using four funnel stands placed at variable distances from naturally BLD-infected beech trees. This approach enabled the recovery of active Lcm nematodes from each funnel, demonstrating their ability to naturally disperse at least 11.74 m from the nearest BLD-infected tree. The findings highlight the role of abiotic factors involved in the dispersal dynamics of Lcm, especially wind and humidity, as indicated by a generalized linear model. The current study also uncovered the incidental association of Lcm with other organisms beneath the canopy of BLD trees, including spiderwebs and caterpillars. To our knowledge, this is the first study to document the potential vectors involved in the local dispersal of Lcm, offering valuable information for the biology of this nematode, as well as insight into the development of effective BLD management strategies. The findings contribute to broader efforts in advancing the understanding of the local spread of BLD, highlighting the complex interplay of abiotic and biotic factors in this disease dispersal.

Evaluation of the impact of wood forest management operations on Amazon nut trees in forest concession areas in the Amazon

The Amazon rainforest holds has immense biodiversity and offers various possibilities for use. Moreover, there is an increasing need for development models that reconcile the rational use of forest resources with socio-economic development. Multiple-use forest management emerges as a legal mechanism for the sustainable use of forest resources. In this context, a forest use system that appears suitable for multiple-use management is one where timber is selectively harvested from forests that are also sources of Brazil nuts. However, the effects of logging management on forests with a high density of Brazil nut trees needs to be studied, as the damage resulting from logging operations on Bertholletia excelsa Bonpl. trees may impact the long-term maintenance of these nut stands. This study aimed to assess the impact of logging operations for timber production on Brazil nut trees to evaluate the compatibility of logging management with the management of native nut stands in a forest concession area in the Amazon. The study was conducted in two National Forests (Jamari and Jacundá) in the state of Rondônia. The damage from cutting and logging operations on the canopy structure of Brazil nut trees was quantified and assessed using photogrammetry with remotely piloted aircraft. The Jamari National Forest showed 4.3% canopy damage to Brazil nut trees, while the Jacundá National Forest showed 3.2% damage. The results indicated that the damage to the canopies of Brazil nut trees due to timber extraction was minimal, suggesting potential compatibility between logging management and the management of native nut stands. However, further studies are needed, especially those focused on planning the collection of Brazil nuts, since access to the nut stands is essential for effectively structuring management, ensuring that the full production potential is sustainably managed and generates income for the extractivists.

Enhanced Satellite Monitoring of Dryland Vegetation Water Potential Through Multi‐Source Sensor Fusion

AbstractDrylands are critical in regulating global carbon sequestration, but the resiliency of these semi‐arid shrub, grassland and forest systems is under threat from global warming and intensifying water stress. We used synergistic satellite optical‐Infrared (IR) and microwave remote sensing observations to quantify plant‐to‐stand level vegetation water potentials and seasonal changes in dryland water stress in the southwestern U.S. Machine‐learning was employed to re‐construct global satellite microwave vegetation optical depth (VOD) retrievals to 500‐m resolution. The re‐constructed results were able to delineate diverse vegetation conditions undetectable from the original 25‐km VOD record, and showed overall favorable correspondence with in situ plant water potential measurements (R from 0.60 to 0.78). The VOD water potential estimates effectively tracked plant water storage changes from hydro‐climate variability over diverse sub‐regions. The re‐constructed VOD record improves satellite capabilities for monitoring the storage and movement of water across the soil‐vegetation‐atmosphere continuum in heterogeneous drylands.

An adaptable dead fuel moisture model for various fuel types and temporal scales tailored for wildfire danger assessment

Estimating the Dead Fuel Moisture Content (DFMC) is crucial in wildfire risk management, representing a key component in forest fire danger rating systems and wildfire simulation models. DFMC fluctuates sub-daily and spatially, influenced by local weather and fuel characteristics. This necessitates models that provide sub-daily fuel moisture conditions for improving wildfire risk management. Many forest fire danger rating systems typically rely on daily fuel moisture models that overlook local fuel characteristics, with consequent impact on wildfire management. The semi-empirical parametric DFMC model proposed addresses these issues by providing hourly dead fuel moisture dynamics, with specific parameters to consider local fuel characteristics. A calibration framework is proposed by adopting Particle Swarm Optimization-type algorithm. In the present study, the calibration framework has been tested by using hourly 10-hours fuel sticks measurements. Implementing this model in forest fire danger rating systems would enhance detail in forest fire danger conditions, advancing wildfire risk management.

A Comprehensive Framework for Understanding Urban Forests as Social-Ecological Systems

AbstractUrban forest management is a multistakeholder, multi-objective situation whereby a surfeit of synergistic or competing goals may exist. Greater research and applied guidance for what works in which urban forest contexts could help improve urban tree and forest outcomes. The challenge in conducting research of this nature is systematic definitions of “what works” and “which contexts” across multidimensional, polycentric urban forest social-ecological systems. This paper presents a comprehensive framework for studying the complexities in urban forest systems (synthesized from numerous other frameworks in the field) that could be used to generate context-specific insights into urban forest management and dynamics. The logic of using frameworks and specific frameworks that already exist within the field are reviewed. Then, I present the urban forest social-ecological system (UFSES) framework. The UFSES framework specifies 5 first-tier factors: theCharacteristics of Trees in the Urban Forest (T); theSurrounding Growing Environment (E);Management & Institutions (M); andCharacteristics of the Human Community (H); which influenceUrban Forest Outcomes (O). A detailed set of second-tier variables nested within these factors are presented in tables at the end of the paper. The framework can foster holistic systems thinking in a systematic yet flexible way; provide a working draft of a common language for thinking about and studying urban forest systems; and enable comparative case research.

The COS CGM Compendium. V. The Dichotomy of O vi Associated with Low- and High-metallicity Cool Gas at z < 1

We analyze the O vi content and kinematics for 126 H i-selected absorbers at 0.14 ≲ z ≲ 0.73 for which the metallicities of their cool photoionized phase have been determined. We separate the absorbers into 100 strong Lyα forest systems (SLFSs with 15 ≲ log N(H i) < 16.2) and 26 partial Lyman Limit systems (pLLSs with 16.2 ≤ log N(H i) ≤ 17.2). The sample is drawn from the COS CGM Compendium (CCC) and has O vi coverage in signal-to-noise ratio ≥ 8 Hubble Space Telescope/COS G130M/G160M QSO spectra, yielding a 2σ completeness level of logN (O vi) ≥ 13.6. The O vi detection rates differ substantially between low-metallicity (LM; [X/H] ≤ −1.4) and high-metallicity (HM; [X/H] > −1.4) SLFSs, with 20% and 60% detection rates, respectively. The O vi detection frequency for the HM and LM pLLSs is, however, similar at ∼60%. The SLFSs and pLLSs without detected O vi are consistent with the absorbing gas being in a single phase, while those with O vi trace multiphase gas. We show that the O vi velocity widths and column densities have different distributions in LM and HM gas. We find a strong correlation between O vi column density and metallicity. The strongest ( logN (O vi) ≳ 14) and broadest O vi absorbers are nearly always associated with HM absorbers, while weaker O vi absorbers are found in both LM and HM absorbers. From comparisons with galaxy-selected and blind O vi surveys, we conclude absorbers with logN (O vi) ≳ 14 most likely arise in the CGM of star-forming galaxies. Absorbers with weak O vi likely trace the extended CGM or intergalactic medium, while those without O vi likely originate in the intergalactic medium.

An integrated dataset of ground hydrothermal regimes and soil nutrients monitored in some previously burned areas in hemiboreal forests in Northeast China during 2016–2022

Abstract. Under a warming climate, the occurrence of wildfires has been becoming increasingly more frequent in boreal forests and Arctic tundra over the last few decades. Wildfires can cause radical changes in forest ecosystems and the permafrost environment, such as the irreversible degradation of permafrost, succession of boreal forests, rapid and massive losses of soil carbon stock, and increased periglacial geohazards. Since 2016, we have gradually and more systematically established a network for studying soil nutrients and monitoring the hydrothermal state of the active layer and near-surface permafrost in the northern Da Xing'anling Mountains in Northeast China. Soil moisture content (depth of 0–9.4 m), soil organic carbon content (0–3.6 m), total nitrogen content (0–3.6 m), and total phosphorus and potassium content (0–3.6 m) datasets were obtained in 2016 via field sampling and subsequent laboratory tests. Ground temperature (0–20 m) and active layer thickness (2017–2022) datasets were obtained using thermistor cables that were permanently installed in boreholes or interpolated with these temperatures. The present data can be used to simulate changes in permafrost features under a changing climate and wildfire disturbances and to explore the changing interactive mechanisms of the fire–permafrost–carbon system in hemiboreal forests. Furthermore, they can provide baseline data for studies and action plans to support the carbon neutralization initiative and assessment of the ecological safety and management of the permafrost environment. These datasets can be easily accessed via the National Tibetan Plateau/Third Pole Environment Data Center (https://doi.org/10.11888/Cryos.tpdc.300933, Li and Jin, 2024).

A practical framework for applied forestry assisted migration

Recent and projected changes in climate over this century pose an unprecedented threat to the health, diversity, and productivity of forest ecosystems. Forests have migrated and adapted to long-term changes in climate over thousands to millions of years; however, natural migration rates and adaptive responses of tree populations cannot match the rapid pace of current climate change. Consequently, more climate-informed approaches to reforestation are needed as current reforestation strategies using local seed sources may no longer be adequate to meet forest management objectives. Assisted migration is a climate change adaptation technique that can help maintain the ecosystem services and economic value that forests provide. Forestry assisted migration (FAM) focusses on the movement of populations of widespread, commercially, or ecologically important forest tree species within or just beyond their current ranges as a way to maintain forest productivity and health in the face of climate change. Although the forestry community recognizes FAM as a reforestation tool, guidance for planning and implementation of FAM is lacking and a framework that provides this guidance can prove useful to land managers with limited time and resources available who want to undertake FAM. We developed a practical framework (the FAM Framework) to provide a structured approach to ensure the most important considerations and best available science are utilized by land managers wanting to implement FAM on their land base. The FAM Framework incorporates multiple factors for the application of FAM in four sequential phases: assessment and analysis, climate-based plant material selection, seed procurement and deployment, and documentation and monitoring. The FAM Framework was tested by developing an assisted migration plan for the Superior National Forest, Minnesota (MN), and lessons learned from the development of this specific plan were used to revise and improve the FAM Framework for suitability across all lands. While originally designed to meet the needs of National Forest System land managers, it is relevant and applicable across the spectrum of land ownership because it incorporates consideration of critical elements in planning and implementing FAM on any landscape while facilitating adaptive management for active learning and future implementation.

Defaunation impacts on the carbon balance of tropical forests.

The urgent need to mitigate and adapt to climate change necessitates a comprehensive understanding of carbon cycling dynamics. Traditionally, global carbon cycle models have focused on vegetation, but recent research suggests that animals can play a significant role in carbon dynamics under some circumstances, potentially enhancing the effectiveness of nature-based solutions to mitigate climate change. However, links between animals, plants, and carbon remain unclear. We explored the complex interactions between defaunation and ecosystem carbon in Earth's most biodiverse and carbon-rich biome, tropical rainforests. Defaunation can change patterns of seed dispersal, granivory, and herbivory in ways that alter tree species composition and, therefore, forest carbon above- and belowground. Most studies we reviewed show that defaunation reduces carbon storage 0-26% in the Neo- and Afrotropics, primarily via population declines in large-seeded, animal-dispersed trees. However, Asian forests are not predicted to experience changes because their high-carbon trees are wind dispersed. Extrapolating these local effects to entire ecosystems implies losses of ∼1.6 Pg CO2 equivalent across the Brazilian Atlantic Forest and 4-9.2 Pg across the Amazon over 100 years and of ∼14.7-26.3 Pg across the Congo basin over 250 years. In addition to being hard to quantify with precision, the effects of defaunation on ecosystem carbon are highly context dependent; outcomes varied based on the balance between antagonist and mutualist species interactions, abiotic conditions, human pressure, and numerous other factors. A combination of experiments, large-scale comparative studies, and mechanistic models could help disentangle the effects of defaunation from other anthropogenic forces in the face of the incredible complexity of tropical forest systems. Overall, our synthesis emphasizes the importance of-and inconsistent results when-integrating animal dynamics into carbon cycle models, which is crucial for developing climate change mitigation strategies and effective policies.

Multiple resiliency metrics reveal complementary drivers of ecosystem persistence: An application to kelp forest systems.

Human-caused global change produces biotic and abiotic conditions that increase the uncertainty and risk of failure of restoration efforts. A focus of managing for resiliency, that is, the ability of the system to respond to disturbance, has the potential to reduce this uncertainty and risk. However, identifying what drives resiliency might depend on how one measures it. An example of a system where identifying how the drivers of different aspects of resiliency can inform restoration under climate change is the northern coast of California, where kelp experienced a decline in coverage of over 95% due to the combination of an intense marine heat wave and the functional extinction of the primary predator of the kelp-grazing purple sea urchin, the sunflower sea star. Although restoration efforts focused on urchin removal and kelp reintroduction in this system are ongoing, the question of how to increase the resiliency of this system to future marine heat waves remains open. In this paper, we introduce a dynamical model that describes a tritrophic food chain of kelp, purple urchins, and a purple urchin predator such as the sunflower sea star. We run a global sensitivity analysis of three different resiliency metrics (recovery likelihood, recovery rate, and resistance to disturbance) of the kelp forest to identify their ecological drivers. We find that each metric depends the most on a unique set of drivers: Recovery likelihood depends the most on live and drift kelp production, recovery rate depends the most on urchin production and feedbacks that determine urchin grazing on live kelp, and resistance depends the most on feedbacks that determine predator consumption of urchins. Therefore, an understanding of the potential role of predator reintroduction or recovery in kelp systems relies on a comprehensive approach to measuring resiliency.

Examination of Exercise Physiological Traits According to Usage Grade of National Forest Trails

Forest trail difficulty levels must be investigated to promote their safe usage. However, exercise-based physiological changes are difficult to estimate in a forest environment. Hence, studies comparing indoor and outdoor environmental activities are required. We aimed to investigate the physiological traits associated with different usage grades of national forest trails in the Republic of Korea. We conducted both field and indoor experiments to measure different variables of 20 healthy participants aged 40–50 years to evaluate the exercise intensity and physiological effects of hiking on these trails. Significant variations in physiological responses were observed based on trail difficulty, with the highest heart rate and energy expenditure recorded during the most challenging uphill segments. Specifically, the heart rate increased from 115.4 ± 13.7 bpm in easy segments to 140.3 ± 15.6 bpm in difficult segments, whereas energy expenditure ranged from 404.1 ± 112.2 kcal/h to 518.1 ± 131.0 kcal/h. This study highlights the importance of considering both objective trail difficulty and users’ physiological reactions, including perceived exertion, to enhance user safety and optimize the health benefits of national forest trails. These findings provide essential data for developing comprehensive exercise programs and improving the national forest trail usage grade system.

Examining the effect of forest certification on timber harvest best management practice implementation

ABSTRACT Forest certification programs are voluntary initiatives emphasizing sustainable forestry and harvest practices. Compliance with certification standards is enforced through third-party audits. One requirement of most forest certification systems is the use of best management practices during harvesting activities. These practices are designed to mitigate potential negative outcomes of forest management and timber harvesting activities. Best management practice implementation has been found to have a variety of positive outcomes on water quality and other resources. However, information relating forest certification to implementation of best management practices and environmental outcomes on certified versus noncertified lands is scant. Here, we examine the role of third-party certification as it contributes to enhanced forest stewardship via implementation of voluntary best management practices. Our study adds to the limited body of research comparing sustainability practices and outcomes on certified vs. noncertified lands. In alignment with previous research on the topic, our study suggests that adherence to best management practices is better on certified versus noncertified forests. Our findings also suggest that for large public agency ownerships, organizationally driven factors beyond forest certification (e.g. institutional policy and trained staff) influence the use of best management practices during timber harvests.

Evaluating soil quality and carbon storage in Western Ghats Forests, Karnataka, India, for sustainable forest management.

Monitoring soil quality index (SQI) and soil organic carbon (SOC) stock status of the Western Ghats (WG) forests in India is crucial for providing vital ecosystem services alongside sustainable forest management practices. However, comprehensive profile data on SQI and SOC stock across different forest types under WG forests are limited. The study evaluated SQI and SOC stock under three forest types, i.e. tropical wet evergreen (TWE), tropical semi-evergreen (TSE), and tropical moist deciduous (TMD) across WG in Karnataka. SQI was assessed using principal component analysis with two indexing approaches and scoring methodologies, with weightage indexing through nonlinear scoring functions (NLSF) showing superiority over other methodologies. TMD forests exhibited the highest SQI, followed by TWE and TSE, while the lowest was observed in Rippon Pet RF (0.36 surface, 0.28 control section), primarily due to limitations in organic carbon and clay content. SOC stock mirrored SQI trends (TMD > TWE > TSE), with the highest values in Kollegal RF (339.3 MG ha-1) and lowest in Rippon Pet RF (102.5 MG ha-1). Although SOC and SQI were established to be ideal indicators for dynamic ecosystem services (ESs), high OC content in surface soils of Poomale NF induces pedogenic acidification and Al toxicities, indicating potential forest soil degradation. Significant correlation with control section SQI and SOC (p < 0.05) emphasises monitoring subsurface soil status to identify soil degradation, sustainable forestry practices, and complex ESs in forest systems.

Harvesting fragmented boreal forest: system selection using a simulation-optimization approach

ABSTRACT Nordic forests, such as those found in Canada were known to offer opportunities for large and relatively homogeneous harvesting blocks. Increased fragmentation of forests makes operation planning more difficult and affects costs of road building and machinery relocation. Currently, the diversity of systems used for forest operations in eastern Canada represents only a fraction of existing alternatives. It might be that at least one alternative outperforms the systems currently used in the fragmented boreal forest. Hence, a subset of potential systems was identified in a preliminary evaluation of harvest systems in fragmented operations. From this sample, the objective is to develop a mathematical model to identify the harvest system with the lowest wood procurement cost in fragmented boreal forests. The candidate systems were simulated and optimized using a static deterministic approach. According to our results, the best systems in fragmented forests involve fewer machines, hence a lower relocation cost. Two CTL systems outperformed all others. The system using removable crane self-loading trucks for transport and the one using the forwarder for loading both resulted in 4 $/m3 (USD) advantage over the third least expensive system in the most fragmented harvest sites. The proposed model can be directly applied to assess harvest systems in other parts of the world with similar forest fragmentation challenges.

Socioecological impacts of pine monocultures on Guaraní territories in Argentina: the hidden costs of modern development

ABSTRACT This study examines the socioenvironmental impacts of pine (Pinus taeda) monocultures established by forestry companies on Guaraní territories in the Province of Misiones, Argentina. Conducted from an interdisciplinary perspective, the research addresses the concerns of a Guaraní community (Puente Quemado II) about the effects on their way of life and the potential toxicological risks resulting from changes to their territory, including large monocultures of pine. By incorporating inputs from ethnography, ethnobiology, biology, ecology, ecotoxicology and toxicology, the study provides a comprehensive analysis of the impacts of the forest monoculture system on the communities. The findings indicate that the implementation of the monoculture system has negative impacts on the communities in terms of cultural and material resources, as well as toxicological risks. Collectively, these factors represent a high risk to their existence within the territory. However, conditions in the area allow rapid ecological restoration, which, with appropriate policies, could serve as a strategy for historical repair efforts and the fulfillmentt of indigenous communities’ rights.

Quercus acutissima exhibits more adaptable water uptake patterns in response to seasonal changes compared to Pinus massoniana

BackgroundSeasonal precipitation variability significantly affects water use in forests; however, whether water uptake is adapted to changes in precipitation, particularly whether it could affect the coexistence of tree species, has rarely been quantified in forest systems. MethodIn this study, dual stable isotopes and the Li-6400 portable photosynthesis system were used to determine the water sources of a mixed conifer (Pinus massoniana) and broadleaf (Quercus acutissima) forest and changes in hydraulic characteristics during the dry and wet seasons in a southern hilly region of China. ResultsAlthough the hydraulic characteristics of P. massoniana were lower than those of Q. acutissima, it maintained a stable water source from the deep soil layer and a higher stomatal conductance (Gs), leading to a higher transpiration rate (Tr) during the growing seasons. Q. acutissima mainly absorbed water from deeper soil layers in the dry season and took up from shallow soil layers in the wet season. Its Gs values exhibited sensitivity to precipitation, while it maintained a lower Tr value during the growing seasons. The excessive water-use strategy observed in P. massoniana may confer weak drought-tolerance during higher frequency and more intense extreme precipitation events, whereas Q. acutissima may exhibit better ecological adaption to precipitation changes. ConclusionsThe overlap of water niches in mixed forests did not appear to affect the coexistence of tree species. The present study provides insights into reforestation and water management in the southern hilly regions of China.

Spatio-temporal dynamic of forest cover distribution in Omo-Shasha-Oluwa Tropical Forest Complex, Nigeria, (2014 – 2022)

: Forest Reserves (FRs) constitute a significant portion of the forested lands within protected areas, and their conservation is essential for Nigeria to achieve the United Nations’ Sustainable Development Goals (SDGs). Our study conducted an analysis of the dynamics of forested lands in the study area to evaluate and rank management performance over the period of 2014-2022. The forest bloc in OSO witnessed an expansion by 1.39% (3,929.67 ha) from 27.73% to 29.12% forest cover in 2014 and 2022, respectively. Forest fragments declined from 16 to 10 units for the forest bloc (&gt;201.1 ha) with a proportionate increase in the Largest Patch Index (LPI) from 5,815.53 ha to 39,234.15 ha, in the period 2014 -2022, respectively. Omo, Ago-Owu, Ife and Shasha FRs all witnessed negative changes in the percent forest cover representing a decline of about -0.43% (529.52 ha), -37.62% (9028.8 ha), -8.54% (1217.43 ha) and -0.29% (89.55 ha), respectively. Oluwa FR had an increase in the forest cover by 17.64% or 14,596.38 ha in the period 2014 - 2022. The forest cover retention rates in the period 2014 – 2022, ranks Oluwa FR ahead of other FRs in terms of Protected Area (PA) management efficiency. The results highlight the existence of a relatively better and effective forest management system in the reserve. Landsat data are capable of providing near real-time forest cover information that can be integrated into existing tracking tools for periodically gauging management efficiency and performances across the PAs of Nigeria.

The Cooling Effect of Oasis Reservoir‐Riparian Forest Systems in Arid Regions

AbstractIn arid regions with limited water resources, numerous reservoirs have been built to support economic and social development. However, how the construction of reservoirs interacts with the surrounding ecosystem to affect temperature remains unclear. Spanning 2018 to 2022 in the Shiyang River Basin, we collected surface water and precipitation, as well as stem and soil samples. Using isotopic methods, we quantified how evaporation in the oasis reservoir‐riparian forest system affects the local climate. Our findings show that the latent heat released by evapotranspiration from the reservoir and riparian forest system reduces the daily maximum temperature and daily temperature range by 7°C and 6°C respectively, compared to downstream areas with sparse vegetation around artificial lakes. Additionally, it enhances local moisture recycling, increasing precipitation. This study reveals regional cooling effect due to interactions between water bodies, the atmosphere, and vegetation. We propose that establishing reservoir‐riparian forest systems can positively impact local climate regulation and serve as an effective strategy for adapting to global climate warming.

Functional traits mediate ant community assembly in a West African savannah-forest mosaic (Côte d'Ivoire)

African forest-savannah mosaics are complex landscapes holding mixtures of woody grasslands (savannah) and different forest systems (gallery forests and forest islands). In these landscapes, ants are highly diverse and perform essential ecosystem services, however, the assembly of ant communities in African forest-savannah mosaics is poorly understood. Here we showed the diversity and species overlap of ant communities in three habitats of the West African savannah and quantified the contribution of thermal tolerances and trophic ecology to community assembly. We investigated ant diversity in the West African Comoé National Park (Côte d'Ivoire) at 16 sites of three habitat types within a forest-savannah mosaic: continuous gallery forest, isolated forest islands and savannah. Across all sites, we collected a total of 91 species from 35 genera from three strata: trees, leaf litter, and soil. Additionally, we assessed differences in functional traits (trophic groups and thermal tolerance) between habitat types and strata. Though species richness was similar in all three habitats, there was a clear separation in species assemblages and functional traits between the two forest habitats and the savannah. Species assemblage shifts were primarily due to species turnover between savannah and forest habitats. In addition, the turnover in species assemblages from forests to savannah habitats was associated with a change in the thermal tolerance of species and in the proportion of trophobionts and predators. Forest and savannah habitats support distinct ant communities with different functional traits and contribute additively to the landscape-scale diversity of the West African ant fauna. Land-use and park management should focus on conserving both savannah and forest sites in tropical protected areas such as the Comoé National Park.

An improved deep convolutional generative adversarial network for quantification of catechins in fermented black tea

The rapid and non-destructive quantification of catechins in fermented black tea is crucial for evaluating the quality of black tea. The combination of hyperspectral imaging and chemometrics has been applied for quantitative detection, but its performance is usually constrained by the limited dataset size. Targeted at the challenge of insufficient samples in regression analysis of catechins, this study proposes an improved deep convolutional generative adversarial network with labeling module, named as DCGAN-L for hyperspectral data augmentation. The DCGAN-L consists of the spectral and label generating modules. First the synthetic spectra were generated, and an indicator was proposed to evaluate their quality. Then, the corresponding label values were generated, including epicatechin gallate (ECG), epicatechin (EGC), catechin (C), and total catechin (CC). For label generating, the Euclidean distances between the synthetic spectrum and all true spectra were measured, followed by allocating weights for calculating the label values based on these distances. Subsequently, the training dataset was augmented with the generated synthetic data. The effect of data augmentation was finally evaluated based on two regression models of random forest (RF) and broad learning system (BLS) for the quantification of catechins. Compared with the results before data augmentation, the average R2 of RF and BLS models increased by 0.044 and 0.164, respectively. The proposed DCGAN-L model allows for the rapid, non-destructive quantification of catechins in black tea in the case of limited sample size.