Response of carbon uptake to abiotic and biotic drivers in an intensively managed Lei bamboo forest

Urbanization expands the fluctuating difference in gross primary productivity between urban and rural areas from 2000 to 2018 in China

Urban and rural vegetation are affected by both climate change and human activities, but the role of urbanization in vegetation productivity is unclear given the dual impacts. Here, we delineated urban area (UA) and rural area (RA), quantified the relative impacts of climate change and human activities on gross primary production (GPP) in 34 major cities (MCs) in China from 2000 to 2018, and analyzed the intrinsic impacts of urbanization on GPP. First, we found that the total urban impervious surface coverage (ISC) of the 34 MCs increased by 13.25 % and the mean annual GPP increased by 211 gC m-2 during the study period. GPP increased significantly in urban core areas, but decreased significantly in urban expansion areas, which was mainly due to a large amount of vegetation loss due to land use conversion. Second, the variability of GPP in UA was generally lower than in RA. Both climate change and human activities had a positive impact on GPP in UA and RA in the 34 MCs, of which the contribution was 49 % and 51 % in UA, and 76 % and 24 % in RA, respectively. Third, under climate change and human activities, the increase in GPP offset 4.96 % and 12.35 % of the impact of land use conversion on GPP in 2000 and 2018, respectively, which indicated that the offset strengthened over time. These findings emphasize the role of human activities in promoting carbon sequestration in urban vegetation, which is crucial for better understanding the processes and mechanisms of urban carbon cycles. Decision-makers can manage urban vegetation based on vegetation carbon sequestration potential as regions urbanize, aiding comprehensive decision-making.

Improving the spatial and temporal estimation of ecosystem respiration using multi-source data and machine learning methods in a rainfed winter wheat cropland

The investigation of ecosystem respiration (RE) and its vital influential factors along with the timely and accurate detection of spatiotemporal variations in RE are essential for guiding agricultural production planning. RE observation in the plot region is primarily based on the laborious chamber method. However, upscaling the spatial-temporal estimates of RE at the canopy scale is still challenging. The present study conducted a field experiment to determine RE using the chamber method. A multi-rotor unmanned aerial vehicle (UAV) equipped with a multispectral camera was employed to acquire the canopy spectral data of wheat during each RE test experiment. Moreover, the agronomic indicators of aboveground plant biomass, leaf area index, leaf dry mass as well as agrometeorological and soil data were measured simultaneously. The study analyzed the potential of multi-information for estimating RE at the field scale and proposed two strategies for RE estimation. In addition, a semiempirical, yet Lloyd and Taylor-based, remote sensing model (LT1-NIR_V) was developed for estimating RE observed across different growth stages with a small margin of error (coefficient of determination [R²] = 0.60-0.64, root-mean-square error [RMSE] = 285.98-316.19 mg m^-2 h^-1). Further, five machine learning (ML) algorithms were utilized to independently estimate RE using two different datasets. The rigorous analyses, which included statistical comparison and cross-validation for estimating RE, confirmed that the XGBoost model, with the highest R² and lowest RMSE (R² = 0.88 and RMSE = 172.70 mg m^-2 h^-1), performed the best among the evaluated ML models. The LT1-NIR_V model was less effective in estimating RE compared with the other ML models. Based on this comprehensive comparison analysis, the ML model can successfully estimate variations in wheat field RE using high-resolution UAV multispectral images and environmental factors from the wheat cropland system, thereby providing a valuable reference for monitoring and upscaling RE observations.

Soil moisture shapes the environmental control mechanism on canopy conductance in a natural oak forest

Canopy conductance (g_c) is an important biophysical parameter closely related to ecosystem energy partitioning and carbon sequestration, which can be used to judge drought effect on forest ecosystems. It is very important to explore how soil moisture change affects the environmental control mechanism of g_c, especially in natural oak forests in Central China where frequent extreme precipitation (P) and drought will occur in a context of climate change. In this study, variations of g_c and its environmental control mechanisms in a warm-temperate forest over three consecutive years under different hydroclimatic conditions were examined by using eddy-covariance technique. Results showed that the averaged g_c in the three growing seasons were 11.2, 11.3 and 7.8 mms^-1, respectively, with a CV of 19.7 %. The lowest g_c occurred in the year with the lowest P. Using three years of data, we found that vapor pressure deficit (VPD) exhibited the dominate effect on g_c, both diffuse photosynthetically active radiation (PAR_dif) and air temperature (T_a) were positively correlated with g_c. When relative extractable water content (REW) was larger than 0.4, however, inhibiting effect of high VPD on g_c disappeared and the effect of direct photosynthetically active radiation (PAR_dir) on g_c was larger compared to PAR_dif. When REW was <0.1, the positive relationship between T_a and g_c became negative. Our results indicated that soil moisture ultimately shapes the environmental control mechanism of g_c in a natural oak forest.

Metabolic Profiling and Transcriptome Analysis Reveal the Key Role of Flavonoids in Internode Coloration of Phyllostachys violascens cv. Viridisulcata

Bamboo, being an ornamental plant, has myriad aesthetic and economic significance. Particularly, Phyllostachys violascens cv. Viridisulcata contains an internode color phenotype in variation in green and yellow color between the sulcus and culm, respectively. This color variation is unique, but the underlying regulatory mechanism is still unknown. In this study, we used metabolomic and transcriptomic strategies to reveal the underlying mechanism of variation in internode color. A total of 81 metabolites were identified, and among those, prunin as a flavanone and rhoifolin as a flavone were discovered at a high level in the culm. We also found 424 differentially expressed genes and investigated three genes (PvGL, PvUF7GT, and PvC12RT1) that might be involved in prunin or rhoifolin biosynthesis. Their validation by qRT-PCR confirmed high transcript levels in the culm. The results revealed that PvGL, PvUF7GT, and PvC12RT1 might promote the accumulation of prunin and rhoifolin which were responsible for the variation in internode color of P. violascens. Our study also provides a glimpse into phenotypic coloration and is also a valuable resource for future studies.

Variations in seasonal and inter-annual carbon fluxes in a semi-arid sandy maize cropland ecosystem in China's Horqin Sandy Land

Sandy cropland ecosystems are major terrestrial ecosystems in semi-arid regions of northern China's Horqin Sandy Land, where they play an important role in the regional carbon balance. Continuous observation of the CO₂ flux was conducted from 2014 to 2018 using the eddy covariance technique in a sandy maize cropland ecosystem in the Horqin Sandy Land. We analyzed carbon fluxes (the net ecosystem exchange (NEE) of CO₂, ecosystem respiration (R_eco), and the gross primary productivity (GPP) and their responses to environmental factors at different temporal scales using Random Forest models and correlation analysis. We found that the sandy cropland was a carbon sink, with an annual mean NEE of -124.4 g C m^-2 yr^-1. However, after accounting for carbon exports and imports, the cropland became a net carbon source, with net biome production ranging from -501.1 to -266.7 g C m^-2 yr^-1. At a daily scale, the Random Forest algorithm revealed that photosynthetic photon flux density, soil temperature, and soil moisture were the main drivers for variation of GPP, R_eco, and NEE at different integration periods. At a monthly scale, GPP and R_eco increased with increasing leaf area index (LAI), so the maize ecosystem's carbon sequestration capacity increased with increasing LAI. At an annual scale, water availability (precipitation and irrigation) played a dominant role in explaining inter-annual variability of GPP and R_eco. Affected by climate (e.g., precipitation) and field management (e.g., cultivation, irrigation), carbon fluxes differed greatly between years in the maize system.

Abiotic and biotic factors contribute to CO2 exchange variation at the hourly scale in a semiarid maize cropland

Understanding the variables influencing the carbon budget in agricultural ecosystems is crucial for the prediction of future carbon dynamics. The purpose of this study was to identify the biotic and abiotic determinants of the net ecosystem CO₂ exchange (NEE) and net assimilation rate (NPP) in a semiarid maize cropland. The CO₂ exchange (NEE and NPP) was measured at different growth stages of maize plants using an improved chamber methodology. Heat map clustering of the correlation coefficients between CO₂ exchange and its driving factors demonstrated that soil temperature and air humidity were positively correlated with CO₂ emissions regardless of daytime or nighttime, while other factors affecting CO₂ exchange were negatively correlated with emissions during daytime yet positively correlated during nighttime. The machine learning algorithm random forest (RF) and structural equation modeling (SEM) were used to analyze the effects of different factors on CO₂ exchange. The RF analysis results indicated that for CO₂ exchange in the daytime, photosynthetically active radiation (PAR) was the most important variable and presented an importance score of 0.574 for NEE and 0.558 for NPP. The SEM results indicated that in the daytime PAR exerted significant direct and indirect effects on both NEE and NPP, and the standardized direct and indirect effects were -0.668 and 0.022, respectively, for NEE, and the effects were 0.655 and -0.011, respectively for NPP. Like PAR, soil water content also exerted significant direct and indirect effects on both NEE and NPP, but the remaining factors affecting CO₂ exchange only have one of the direct or indirect effects, sometimes neither. For CO₂ exchange at night, the leaf area was the most important variable and presented an importance score of 0.72 for NEE and 0.45 for NPP. At night, both the direct and indirect effects of most abiotic factors on NEE and NPP were significant.

Predicting carbon and water vapor fluxes using machine learning and novel feature ranking algorithms

Gap-filling eddy covariance flux data using quantitative approaches has increased over the past decade. Numerous methods have been proposed previously, including look-up table approaches, parametric methods, process-based models, and machine learning. Particularly, the REddyProc package from the Max Planck Institute for Biogeochemistry and ONEFlux package from AmeriFlux have been widely used in many studies. However, there is no consensus regarding the optimal model and feature selection method that could be used for predicting different flux targets (Net Ecosystem Exchange, NEE; or Evapotranspiration -ET), due to the limited systematic comparative research based on the identical site-data. Here, we compared NEE and ET gap-filling/prediction performance of the least-square-based linear model, artificial neural network, random forest (RF), and support vector machine (SVM) using data obtained from four major row-crop and forage agroecosystems located in the subtropical or the climate-transition zones in the US. Additionally, we tested the impacts of different training-testing data partitioning settings, including a 10-fold time-series sequential (10FTS), a 10-fold cross validation (CV) routine with single data point (10FCV), daily (10FCVD), weekly (10FCVW) and monthly (10FCVM) gap length, and a 7/14-day flanking window (FW) approach; and implemented a novel Sliced Inverse Regression-based Recursive Feature Elimination algorithm (SIRRFE). We benchmarked the model performance against REddyProc and ONEFlux-produced results. Our results indicated that accurate NEE and ET prediction models could be systematically constructed using SVM/RF and only a few top informative features. The gap-filling performance of ONEFlux is generally satisfactory (R² = 0.39-0.71), but results from REddyProc could be very limited or even unreliable in many cases (R² = 0.01-0.67). Overall, SIRRFE-refined SVM models yielded excellent results for predicting NEE (R² = 0.46-0.92) and ET (R² = 0.74-0.91). Finally, the performance of various models was greatly affected by the types of ecosystem, predicting targets, and training algorithms; but was insensitive towards training-testing partitioning. Our research provided more insights into constructing novel gap-filling models and understanding the underlying drivers affecting boundary layer carbon/water fluxes on an ecosystem level.

Comparative Analysis of Two Machine Learning Algorithms in Predicting Site-Level Net Ecosystem Exchange in Major Biomes

The net ecosystem CO2 exchange (NEE) is a critical parameter for quantifying terrestrial ecosystems and their contributions to the ongoing climate change. The accumulation of ecological data is calling for more advanced quantitative approaches for assisting NEE prediction. In this study, we applied two widely used machine learning algorithms, Random Forest (RF) and Extreme Gradient Boosting (XGBoost), to build models for simulating NEE in major biomes based on the FLUXNET dataset. Both models accurately predicted NEE in all biomes, while XGBoost had higher computational efficiency (6~62 times faster than RF). Among environmental variables, net solar radiation, soil water content, and soil temperature are the most important variables, while precipitation and wind speed are less important variables in simulating temporal variations of site-level NEE as shown by both models. Both models perform consistently well for extreme climate conditions. Extreme heat and dryness led to much worse model performance in grassland (extreme heat: R2 = 0.66~0.71, normal: R2 = 0.78~0.81; extreme dryness: R2 = 0.14~0.30, normal: R2 = 0.54~0.55), but the impact on forest is less (extreme heat: R2 = 0.50~0.78, normal: R2 = 0.59~0.87; extreme dryness: R2 = 0.86~0.90, normal: R2 = 0.81~0.85). Extreme wet condition did not change model performance in forest ecosystems (with R2 changing −0.03~0.03 compared with normal) but led to substantial reduction in model performance in cropland (with R2 decreasing 0.20~0.27 compared with normal). Extreme cold condition did not lead to much changes in model performance in forest and woody savannas (with R2 decreasing 0.01~0.08 and 0.09 compared with normal, respectively). Our study showed that both models need training samples at daily timesteps of >2.5 years to reach a good model performance and >5.4 years of daily samples to reach an optimal model performance. In summary, both RF and XGBoost are applicable machine learning algorithms for predicting ecosystem NEE, and XGBoost algorithm is more feasible than RF in terms of accuracy and efficiency.

Carbon fluxes response of an artificial sand-binding vegetation system to rainfall variation during the growing season in the Tengger Desert

Revegetation is considered as an effective approach for desertification control, and artificial sand-binding vegetation exerts a significant contributor to carbon cycling in arid and semiarid regions; however, this is largely determined by the rainfall regime. We measured carbon fluxes (the net ecosystem CO₂ exchange (NEE), the gross ecosystem productivity (GEP) and the ecosystem respiration (R_eco)) during the growing season of 2014-2016 using the eddy covariance technique and explored the effects of rainfall variables (amount, timing distribution and pulse size) and environmental factors on carbon fluxes at different time scales. The system had NEE values of -117.5 and -98.9 g C m^-2 during the growing seasons of 2015 (dry year) and 2016 (wet year), respectively. When the rainfall amount did not differ significantly between spring and autumn, the cumulative GEP was greater in spring than in autumn, whereas the cumulative R_eco and NEE showed the opposite pattern. Small (<5 mm) rain events failed to trigger obvious GEP and NEE pulses, whereas ≥ 5 mm or a series of small rain events led to high assimilation but with hysteresis. The magnitude of R_eco increased as the rain pulses increased. The Random Forest (RF) algorithm revealed that soil water contents had a great impact on carbon fluxes at different integration periods. A correlation analysis showed that the soil water contents were positively correlated with GEP and R_eco and negatively correlated with NEE over different time scales in most cases. These findings suggest that artificial vegetation not only improves habitat restoration but is a significant carbon sink during both dry and wet growing season, which is likely to supplement our knowledge gap to accurately evaluate the current carbon budget in dry land.

Response of ecosystem functioning to environmental variations in an artificial sand-binding vegetation desert in northwestern China

The establishment of artificial sand-binding vegetation is one of the main means for restoring damaged ecosystems that are impacted by global change. This study was conducted to evaluate the influence of environmental factors on ecosystem function (net ecosystem exchange (NEE), evapotranspiration (ET), and water use efficiency (WUE)) in an artificial sand-binding vegetation desert (with dominant shrubs, such as Artemisia ordosica and Caragana korshinskii, and herbaceous plants) in northwestern China. NEE, ET, and meteorological data were observed with the eddy covariance (EC) technique. The random forest (RF) method was used to identify major environmental factors that affected NEE, ET, and WUE. Our results showed that the mean annual NEE, ET, and WUE values were - 112.4 g C m^-2, 232.1 mm, and 0.49 g C kg^-1 H₂O, respectively, during the 2015 to 2018 growing seasons. At the weekly scale, the most important drivers of NEE were the normalized difference vegetation index (NDVI) and soil water content (SWC). Rainfall, SWC, and NDVI were important drivers of ET. WUE was mainly controlled by rainfall and SWC. Linear regression showed that NEE had significant negative relationships with the NDVI and SWC. ET had positive relationships with rainfall, SWC, and the NDVI. WUE had significant negative relationships with SWC and rainfall. These findings indicate that drought inhibited ET more than carbon absorption, thus promoting the WUE of the ecosystem to some extent. The close relation of the ecosystem function to SWC implies that this ecosystem may be critically regulated by future climate change (specifically, changes in rainfall patterns).

Changes of Root Endophytic Bacterial Community Along a Chronosequence of Intensively Managed Lei Bamboo (Phyllostachys praecox) Forests in Subtropical China

Endophytic bacteria widely exist inside plant tissues and have an important role in plant growth and development and the alleviation of environmental stress. However, little is known about the response of root-associated bacterial endophytes of Lei bamboo (Phyllostachys praecox) to intensive management, which is a common management practice for high bamboo shoot production in subtropical China. In this study, we comparatively investigated the root endophytic bacterial community structures in a chronosequence of intensively managed (5a, 10a, 15a, and 20a) and extensively managed plantations (as control, Con). The results showed that endophytic Proteobacteria was the dominant bacterial phylum in the bamboo roots. Intensive management significantly increased (p < 0.05) the bacterial observed species and Chao1 (except 5a) indices associated with bamboo roots. The relative abundances of Firmicutes, Bacteroidetes, and Actinobacteria (except 15a) in the intensively managed bamboo roots significantly increased (p < 0.05) compared with those in Con, while the relative abundance of Proteobacteria significantly decreased in intensively managed bamboo roots (p < 0.05). The phyla Proteobacteria, Actinobacteria, Bacteroidetes, and Firmicutes were the biomarkers in Con, 5a, 15a, and 20a, respectively. Redundancy analysis (RDA) showed that soil alkali-hydrolysable N (AN), available phosphorus (AP), available K (AK), and total organic carbon (TOC) were significantly correlated (p < 0.05) with the bacterial community compositions. Our results suggest that the root endophytic microbiome of Lei bamboo was markedly influenced by intensive management practices, and the available nutrient status could be the main driving factor for such shifts. Although heavy fertilization in the intensive management system increased the diversity indices, the rapid changes in root endophyte communities and their relevant functions might indicate a high risk for sustainable management.

Spatiotemporal evolution and impacts of climate change on bamboo distribution in China

Understanding the impact and restriction of climate change on potential distribution of bamboo forest is crucial for sustainable management of bamboo forest and bamboo-based economic development. In this study, climatic variables and maximum entropy model were used to simulate the potential distribution of bamboo forest in China under the future climate scenarios. Seven climatic variables, such as Spring precipitation, Summer precipitation, Autumn precipitation, average annual relative humidity, Autumn average temperature, average annual temperature range and annual total radiation, were selected as input variables of maximum entropy model based on the relative importance of those climate variables for predicting bamboo forest presence. The suitable ranges of the seven climatic variables for potential distribution of bamboo forest were 337-794 mm, 496-705 mm, 213-929 mm, 74.3%-83.4%, 16.6-23.8 °C, 2.3-10.1 °C and 3.2 × 10⁴-4.3 × 10⁴ W m^-2, respectively. Under RCP4.5 and RCP8.5 climate scenarios, the suitable area of bamboo forest growth first increased and then decreased, and showed range contractions towards the interior and expansions towards southwest in China. The results of the present study can serve as a useful reference to dynamic monitoring of the spatial distribution and sustainable utilization of bamboo forest in the future under climate change.

Investigation on the carbon sequestration capacity of vegetation along a heavy traffic load expressway

Carbon sequestration by vegetation plays an important role in the global carbon cycle. More emphasis on the carbon sequestration of roadside vegetation will help to reduce the total carbon emissions from the transportation sector. In the current study, the Shanghai-Nanjing G42 expressway in east China was selected to investigate and calculate the carbon sequestration of roadside vegetation including trees, shrubs and herbs. Findings of the current study revealed that the total carbon sequestration of all the vegetation was about 97,000 tons per year. Results also indicated that trees have a higher carbon sequestration capacity (γ) in unit land area compared to shrubs and herbs. The γ value of most of the shrubs was lower than that of tree; however, species such as Nerium indicum, Jasminum mesnyi and Forsythia suspense have better carbon sequestration capacity than some other tree species. The γ value of herbs was too low, compared with trees and shrubs. The findings of the current study will be of great benefit to make the vegetation planting strategy for express highways in the areas with similar geographic characteristics and climate.

Water-Use Characteristics and Physiological Response of Moso Bamboo to Flash Droughts

Frequent flash droughts can rapidly lead to water shortage, which affects the stability of ecosystems. This study determines the water-use characteristics and physiological mechanisms underlying Moso bamboo response to flash-drought events, and estimates changes to water budgets caused by extreme drought. We analyzed the variability in forest canopy transpiration versus precipitation from 2011-2013. Evapotranspiration reached 730 mm during flash drought years. When the vapor pressure deficit > 2 kPa and evapotranspiration > 4.27 mm·day^-1, evapotranspiration was mainly controlled through stomatal opening and closing to reduce water loss. However, water exchange mainly occurred in the upper 0-50 cm of the soil. When soil volumetric water content of 50 cm was lower than 0.17 m³·m^-3, physiological dehydration occurred in Moso bamboo to reduce transpiration by defoliation, which leads to water-use efficiency decrease. When mean stand density was <3500 trees·ha^-1, the bamboo forest can safely survive the flash drought. Therefore, we recommend thinning Moso bamboo as a management strategy to reduce transpiration in response to future extreme drought events. Additionally, the response function of soil volumetric water content should be used to better simulate evapotranspiration, especially when soil water is limited.