Woody component of tropical rainforest recovers slower from drought than the upper canopy layer and leaves

Woody component of tropical rainforest recovers slower from drought than the upper canopy layer and leaves | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Woody component of tropical rainforest recovers slower from drought than the upper canopy layer and leaves Feng Tian, Yujie Dou, Jean-Pierre Wigneron, Xiaojun Li, Wenmin Zhang, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4464016/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 20 Nov, 2024 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Abstract Tropical rainforests are crucial for Earth's health, but climate change is making severe droughts more frequent. The 2015–2016 El Niño-induced drought caused significant biomass loss, yet the recovery duration of different vegetation components (woody parts, upper canopies, and leaves) remains unknown. This study employed satellite remote sensing data of L-band Vegetation Optical Depth (L-VOD), X-band VOD (X-VOD), and Enhanced Vegetation Index (EVI) from 2010 to 2022, characterized by having different sensitivities to the different vegetation components, to examine the recovery of these components in the tropical evergreen broadleaf forest (EBF) regions during the 2015–2016 El Niño-induced drought. Results showed that the woody component had the slowest recovery, particularly in Africa, which took longer to return to pre-drought conditions than South America. Key factors influencing recovery included drought severity, moisture-related climatic conditions (i.e., VPD, precipitation, and soil moisture), and seasonal variations. Moreover, the woody component of the EBF in South America showed less impact from drought, benefitted from more favorable moisture-related climatic conditions (e.g., more precipitation and lower VPD), and experienced higher seasonal variation in monthly temperature and precipitation, resulting in a faster recovery than that observed in Africa. Drought Vegetation recovery Tropical forest Vegetation optical depth Figures Figure 1 Figure 2 Figure 3 Figure 4 1 Introduction Tropical ecosystems represent 34% of the global gross primary terrestrial productivity (Beer et al., 2010 ) and play a major role in carbon cycles at the global scale (Bonal et al., 2016 ). However, the effectiveness of capturing and storing carbon to mitigate future global warming partly depends on the impact of severe drought episodes as water is the primary determinant of the amount and allocation of forest biomass production, and thereby the interannual variability of the tropical carbon cycle (He et al., 2023 ). Droughts in tropical regions are predominantly associated with the El Niño-Southern Oscillation (El Niño), and many extreme drought events in tropical regions coincide with El Niño events (Marengo et al., 2011 ). Notably the 2015–2016 El Niño led to historically high temperatures and low precipitation across the tropics, and the growth rate of atmospheric carbon dioxide was the largest on record (Liu et al., 2017 ). An earlier study found that the carbon stocks in African and American humid forests had not recovered to pre- El Niño levels by 2017 (Wigneron et al., 2020 ), and the duration of the vegetation recovery period has yet to be determined. Different vegetation components are characterized by differences in response time during drought conditions. Several experiments have demonstrated that the sensitivity of woody growth rate to drought surpasses that of vegetation canopy greenness (Gazol et al., 2018 ) because vegetation growth reduction is more mediated by the functional processes related to building a carbon sink than by the quantity of biomass synthesized through photosynthesis (Sarris et al., 2007 ; Zhang et al., 2018 ). However, these experiments were conducted at the species level, and the spatial variability in the sensitivity of different parts of woody plants remains unexplored. The response of forests to drought does not only depend on forest resistance and adaptation strategies, but is also highly dependent on the severity of drought events (Taeger et al., 2013 ), the time scale at which drought occurs (Hahn et al., 2021 ), and the duration of the drought. For example, larger resistance to drought has been observed in spring when vegetation is in its reproductive stage, and productivity is at its peak (Hahn et al., 2021 ). Tropical tall forests are found to be more sensitive and vulnerable to drought than short forests ( Liu et al., 2022 ), and a higher species diversity could enhance drought resistance (Anderegg, Konings, et al., 2018 ). Therefore, to better understand drought impacts on tropical forest ecosystems, it is necessary to consider and incorporate information on forest traits, drought severity, and climate to investigate the potential drivers of forest recovery from drought. Since microwave observations at different frequencies can penetrate the vegetation layer at different depths, the microwave vegetation optical depth (VOD) data derived from multi-frequency microwave spaceborne observations provide a new way to investigate the sensitivity of vegetation to drought across various tree components. Lower frequencies (i.e. longer wavelengths) have deeper penetration depth because low-frequency radiation is less extinguished within the canopy. The L-band VOD (L-VOD) (Wigneron et al., 2024 ) has been demonstrated to convey essential information regarding the larger branches and trunks of woody components (Tian et al., 2017 ), while X-band VOD (X-VOD) is indicative of small branches and leaves of the upper canopy layer (Frappart et al., 2020 ; Li et al., 2021 ). Additionally, the Enhanced Vegetation Index (EVI) derived from optical satellite data is sensitive to vegetation greenness and can be considered a representation of leaf photosynthetic activity (Gerard et al., 2020 ). Consequently, the information captured by L-VOD, X-VOD, and EVI, respectively, enables an exploration of the sensitivity of distinct vertical structures of the vegetation to drought conditions. In this study, we used satellite remote sensing data of L-VOD, X-VOD, and EVI as proxies for the woody component including branches and trunks, the upper layer of canopies, and the leaf component of tropical forest trees, respectively. To understand their respective sensitivity to drought, we aimed to investigate the recovery time of the different components of tropical evergreen forest (EBF) across the pan-tropics following the El Niño-induced drought of 2015–2016, by analyzing time series of satellite data from 2010–2022. We also employed the random forest method to ancillary data of climatic conditions, drought-related information, and ecosystem-related factors to investigate the primary drivers of spatial variability in the recovery time of the woody component in tropical EBF. 2 Data and methods 2.1 Vegetation indices data The SMOS-SMAP-INRAE-BORDEAUX (SMOSMAP-IB) L-VOD product was retrieved from temperature brightness observed from ESA’s Soil Moisture Ocean Salinity (SMOS) and NASA’s Soil Moisture Active Passive (SMAP). It offers L-VOD at a semi-daily temporal resolution and a grid resolution of 25 km (Li et al., 2022 ). The semi-daily global LPDR X-VOD dataset at 0.25 \(^\circ\) spatial resolution, derived from AMSR-E (Advanced Microwave Scanning Radiometer – Earth Observing System) and AMSR-2 (Advanced Microwave Scanning Radiometer − 2) sensors (Du et al., 2017 ) was also employed. Both nighttime L-VOD (ascending orbit, 6:00 AM) and X-VOD (descending orbit, 1:30 AM) were aggregated to monthly data by averaging, covering the period from 2010 to 2022. The MODIS EVI data Version 6.1 from 2010 to 2022 was used, with monthly EVI data aggregated by taking the maximum values. All the vegetation index data have been converted to a geographic coordinate system format (i.e., WGS1984), with a spatial resolution of 0.25°. 2.2 Drought data We used the Standardized Precipitation Evapotranspiration Index (SPEI) as a drought indicator to assess the emergence, length, and intensity of drought events. As the humid forest has been shown to respond to drought within three months (Vicente-Serrano et al., 2013 ), the SPEI03 data with 0.05° spatial resolution (Gebrechorkos et al., 2023 ) was selected, which was calculated from Multi-Source Weighted-Ensemble Precipitation (MSWEP) and potential evapotranspiration (PET) from the Global Land Evaporation Amsterdam Model (GLEAM). The SPEI03 is calculated by factoring in the past 3-month aggregated precipitation and potential evapotranspiration, thus reflecting relatively short-term moisture conditions (Vicente-Serrano et al., 2010 ). The SPEI data were resampled to the spatial resolution of VOD data at 0.25° by averaging. 2.3 Ancillary data Monthly air temperature, dewpoint temperature, and soil moisture data at 0.1° resolution from 2010 to 2022 were taken from the ERA5 monthly average reanalysis dataset. The soil moisture data covers three layers, including layer 1 (0–7 cm), layer 2 (7–28 cm), and layer 3 (28–100 cm). The air temperature and dewpoint temperature were used to calculate the vapor pressure deficit (VPD) using the method provided by Yuan et al., ( 2019 ). The precipitation data was derived from the MSWEP product with a 3 hours temporal resolution and 0.1° spatial resolution from 1979 to the present (Beck et al., 2019 ). All the climate data mentioned above were resampled to the spatial resolution of VOD data by averaging. We included the “elasticity of substitution” data, which reflects the degree to which various species can substitute each other in enhancing forest productivity (Liang et al., 2016 ), the magnitude of the intrinsic variability of vegetation water content data to represent vegetation water buffering (Dou et al., 2023 ), and a proxy for vegetation biomass data (i.e. the mean annual L-VOD pre- El Niño year). Note that the mean annual VOD before El Niño was computed as the 95th percentile of nighttime VOD from 2010 to 2014 as recommended by Dou et al., ( 2023 ). MODIS landcover Version 6.1 data were used to delineate the EBF regions. Areas within VOD footprints where urban and cropland cover exceeds 5% were masked. Additionally, considering the potential impact of urban and cropland expansion on vegetation, we also excluded pixels with proportions of urban and cropland expansion exceeding 5% and pixels with forest loss exceeding 5% based on Hansen’s global forest change data (Hansen et al., 2013 ). 2.4 Calculation of the recovery time Monthly SPEI and vegetation indices (VIs, i.e. L-VOD, X-VOD, and EVI) were used together to identify drought events and the calculation of recovery time for different vegetation components at pixel-scale (Schwalm et al., 2017 ; Yao et al., 2023 ). The monthly VIs time-series data were deseasonalized by subtracting the monthly average values (calculated from the full period excluding the drought years from 2015 to 2016) from the VIs time series to remove the effects of the seasonal cycle and then detrended to eliminate the long-term trend. When a drought event happens and the standardized deviation (SD) of the detrended VI data falls below − 0.5 SD, the vegetation is considered to have been negatively affected by the drought event (Yao et al., 2023 ). The drought event was considered to begin when the SPEI was lower than − 1 and the detrended VI data at the same time were below − 0.5 SD, and it ended when the SPEI was higher than − 1 and the detrended VIs data remained below − 0.5 SD, or SPEI stayed lower than − 1 but the detrended VIs data increased above − 0.5 SD. Additionally, we only focused on drought events lasting for at least 2 months. The calculation of the drought recovery time for different vegetation components is based on the following criteria (Yao et al., 2023 ): (Yao et al., 2023 ): (1) If the detrended VI data reached a local minimum during the drought period (see above), the recovery time was defined as the period from the time when the detrended VI data reached the minimum value to the time when the detrended VI data were higher than − 0.5 SD. (2) If the condition above was not met, the recovery time was defined as the period from the end of the drought event (see above) to the time when the detrended VI data were larger than − 0.5 SD. 2.5 Drivers of the recovery of vegetation woody component The leaf component generally represents only a small fraction of the entire above-ground woody biomass and may not be representative of the trends and spatial patterns of the woody component (Tian et al., 2017 ). Thus, only the drivers of the recovery of vegetation woody components were investigated. We included 44 response variables in the random forest regression model to investigate the relative significance of these variables to the recovery time of the woody component. The 44 response variables were reclassified into four classes, including the normal climatic conditions, the climatic conditions during the recovery period, drought-related factors, and ecosystem-related factors. Specifically, the variables covering the normal climatic conditions comprise annual mean, 25th percentile minimum, 75th percentile maximum, and standard deviation of air temperatures (T_mean, T_min25, T_max75, T_std), precipitation (P_mean, P_min25, P_max75, P_std), soil moisture from layers 1 to 3 (SM1_mean, SM1_min25, SM1_max75, SM1_std, SM2_mean, SM2_min25, SM2_max75, SM2_std, SM3_mean, SM3_min25, SM3_max75, SM3_std), and VPD (VPD_mean, VPD_min25, VPD_max75, VPD_std), covering the period from 2010 to 2022, excluding 2015 to 2016. The standard deviation of air temperatures, precipitation, soil moisture, and VPD were defined as the seasonal variation of normal climate conditions in this study. The variables denoting the climatic conditions during the recovery period include monthly mean air temperature (Recovery_T_mean), precipitation (Recovery_P_mean), VPD (Recovery_VPD_mean), and soil moisture from layers 1 to 3 (Recovery_SM1_mean, Recovery_SM2_mean, Recovery_SM3_mean). Drought-related factors encompass drought duration, drought severity (i.e. mean SPEI), the number of dry months (monthly precipitation less than 100 mm), anomalies of the climatic conditions relative to the pre-El Niño period, including temperature (T_anomaly), precipitation (P_anomaly), VPD (VPD_anomaly), and soil moisture from layer1 to layer3 (SM1_anomaly, SM2_anomaly, SM3_anomaly), and the severity of the drought impact on the woody component (i.e. the L-VOD anomaly during the drought period relative to the pre-drought condition, L-VOD_anomaly). Ecosystem-related variables include the magnitude of the intrinsic variability of vegetation water content data representing vegetation water buffering (Mean_delta_VOD_day), biomass (i.e. pre-El Niño annual VOD values, annual L-VOD), and the elasticity of substitution. The random forest regression model can explain interactions and nonlinear relationships between predictors (Breiman, 2001 ). The importance of each response variable was assessed through the percentage increase in the mean square error (%IncMSE) between target and response values (Delgado-Baquerizo et al., 2017 ). The values of the %IncMSE were generated from a random forest model consisting of 500 decision trees in this study, and higher values of %IncMSE suggest higher importance of the response variables. It should be noted that random forest is a tree-based ensemble model that is sensitive to the nonlinear relationships and interactions among features. Therefore, even if a feature shows high importance in terms of "%IncMSE", its coefficient of determination between recovery time and the response variables may not necessarily be high (Liu et al., 2023 ). Additionally, we conducted a principal component analysis (PCA) on the input data to transform highly correlated input factors (e.g., precipitation and VPD) into a set of uncorrelated principal components to mitigate the impact of collinearity. These principal components were used as the new input features for training the random forest model. 3 Results 3.1 The severity of drought Most EBF in South America (56%) and Africa (90%) regions have experienced severe drought (i.e., SPEI < -1.5) (Fig. 1 a) caused by the 2015–2016 El Niño, and there were obvious differences in the drought duration across the pantropical area (Fig. 1 b). The EBF in Africa showed the most widespread exposure to long-duration droughts, with drought periods lasting up to 6 months covering 93% of the region, whereas 58% of South America forests and 40% of Asian forests have been exposed to such long-duration droughts, respectively. 3.2 Recovery time of different vegetation components Noticeable spatial differences in the recovery time of the woody component (branches and woody), the upper canopy layer, and the leaf component were observed following the 2015–2016 drought (Figure. 2). The recovery of the upper canopy layer and leaves was faster than the woody component. Nearly 73% of the upper canopy layer area and 88% of the leaf area recovered to the pre- El Niño conditions within two months, while only 52% of the drought-affected woody component area showed a similar recovery time (Fig. S1 a). Moreover, there were 29% of the area for the woody component did not recover within one year (Fig. S1 a). with 73% of the area having recovered within two months for the upper canopy layer and 88% in the case of leaves. The recovery time of the upper canopy layer and the leaf component exhibited less spatial variation, but there were notable variations in the recovery time of the woody components in the forest regions of South America, Africa, and Asia. The woody components of EBF in South America show the fastest recovery (Fig. 2 a), with 57% of the region recovering within 2 months. The recovery time for EBF in Africa is the longest, with 42% of the region requiring more than 12 months to fully recover. As for the recovery time for different vegetation components, the pixel scale leaf component recovers first, followed by the upper canopy layer and the woody components (Fig. 2 , Fig. S1 -S2). In 89% of forested areas, the leaf component showed simultaneous recovery time with either the upper canopy layer, the woody component, or both (in 12% of the study area, the leaf component recovered first preceding the recovery of the upper canopy layer and the woody component) (Fig. 2 b). This phenomenon was common in South America, Africa, and Asia (Fig. S2). The regions where the upper canopy layer showed simultaneous recovery time with either the leaf component, the woody component, or both account for 74% (in 4% of forested areas, the upper canopy layer recovered first preceding the recovery of the other two components). Nearly 20% of forested areas showed simultaneous recovery time of the woody component with either the leaf component, the upper canopy layer, or both (the areas where the woody component recovered first accounts for 2%). 3.3 Main drivers of woody component recovery across global tropical rainforests The recovery time of the upper canopy layer and the leaf component exhibited less spatial variation in recovery time (Fig. 2 ), so we solely investigated the factors influencing the recovery time of the woody component. The input factors incorporated in the random forest model collectively explained 70% of the variability in the woody component recovery across the tropical region of EBF forests. The severity of the drought impact on the woody component (i.e. L-VOD anomaly) emerged as the primary determinant of the recovery time (%IncMSE = 102%). Moreover, there was a significant negative correlation between L-VOD anomalies and the recovery time of the woody component (R = -0.58, p < 0.01), indicating that more severe impacts of drought on the woody component result in longer recovery time (Fig. 3 b). Subsequently, the climatic conditions related to moisture levels during the recovery period, including mean monthly VPD (%IncMSE = 32%, R = -0.13, p < 0.01) and mean monthly precipitation (%IncMSE = 29%, R = 0.03, p < 0.05), were found to be important explanatory variables of the recovery time. Biomass (i.e. annual L-VOD) is also closely related to the recovery of vegetation woody components (%IncMSE = 26%, R = 0.27, p < 0.01). The higher the vegetation biomass, the longer the corresponding recovery time. The seasonal variation (i.e., the standard deviation of monthly air temperature and precipitation) also played an important role in influencing the recovery of the vegetation woody component. When the seasonal showed minor variations (e.g., smaller standard deviations in monthly temperature (%IncMSE = 32%, R = -0.24, p < 0.01) and monthly precipitation (%IncMSE = 32%, R = -0.29, p 0.05), soil moisture of layer 3 during the recovery period (%IncMSE = 19%, R = -0.03, p > 0.05), and the long-term mean monthly precipitation (%IncMSE = 19%, R = -0.12, p < 0.01) also positively contributed to the recovery of the woody component, with more water availability corresponded to shorter recovery time. We also analyzed the primary influencing factors on the recovery time of EBF in South America (56% explanation), Africa (73% explanation), and Asia (54% explanation) separately (Fig. S3 - S5). We found that the severity of the drought impact on the woody component (i.e. L-VOD anomaly) consistently remained the most significant factor affecting the recovery time of the woody component over all three continents. This was followed by the climatic conditions during the recovery period in South America and Africa, as well as the drought-related factors (e.g. air temperature anomaly, VPD anomaly) in Asia. 3.4 Possible explanations for the faster recovery of the woody component in South America compared to Africa The primary drivers of the woody components recovery in rainforests worldwide could well explain why these components recovered more quickly in South America than in Africa. Firstly, the severity of the drought impact on the woody component (i.e. L-VOD anomaly) was more pronounced in Africa than in South America (mean L-VOD anomaly = -3.0 SD in Africa vs . -2.3 SD in South America) (Fig. 3 a). In Africa, the most negative impact of drought on the woody component was concentrated in the Congo Rainforest region, where the L-VOD anomaly was lower than in other regions (Fig. S6). Secondly, the climatic conditions related to moisture levels during the recovery period in South American EBF regions were more favorable than in Africa with lower VPD (mean VPD = 5.9 in South America vs . 6.3 in Africa), more precipitation (mean precipitation = 182 mm in South America vs . 136 mm in Africa), and more soil moisture (mean soil moisture of layer 2 = 0.39 m 3 m − 3 in South America vs . 0.37 m 3 m − 3 in Africa, mean soil moisture of layer 3 = 0.39 m 3 m − 3 in South America vs . 0.35 m 3 m − 3 in Africa) (Fig. 4 b-e). The pre-El Niño annual VOD in EBF of South America was slightly higher than in Africa (mean pre-El Niño annual VOD = 0.91 in South America vs . 0.87 in Africa) (Fig. 4 f), but the impact of drought on the woody component and the influence of climate on the recovery period is greater than that of biomass alone (Fig. 3 a). Therefore, solely comparing biomass values cannot adequately explain why the recovery time of the woody vegetation layer in South America is faster than that in Africa. Finally, the EBF in Africa showed a more stable seasonal variation with lower monthly temperature variation (mean temperature variation = 0.9 SD in South America vs . 0.88 SD in Africa) and monthly precipitation variation (mean precipitation variation = 106 SD in South America vs . 71 SD in Africa). Moreover, the EBF has lower precipitation availability in Africa than so (mean precipitation = 207 mm in South America vs. 137 mm in Africa). Therefore, the severity of drought impact on the woody component, the less favorable moisture-related climatic conditions during recovery, and the smaller seasonal variation in Africa compared to South America's EBF contribute to a slower recovery of the woody component in Africa compared to South America. It should be noted that due to the relatively smaller area of EBF in Asia compared to South America and Africa, we chose only to focus our analysis on understanding the differences in the recovery time of woody components in vegetation between South America and Africa. 4 Discussion In this study, we used L-VOD, X-VOD, and EVI data separately as proxies of the woody component (branches and woody), the upper canopy layer, and the leaf component to investigate the recovery of these components in tropical EBF regions during the extreme drought period induced by the 2015–2016 El Niño event. We found that the recovery time of the leaf component is the fastest, followed by the upper canopy layer and the woody component (Fig. 1 , Fig. S1 -S2). During a drought period, the leaf component initially responds by closing stomata, followed by a reaction in the woody parts (Loewenstein & Pallardy, 1998 ). While the longstanding theory holds that stomata optimize fitness by maintaining constant marginal water use efficiency over a specified time frame, a recent evolutionary theory suggests an alternative perspective that stomata aim to maximize the carbon gain while minimizing carbon costs and the risk of hydraulic damage (Anderegg, Wolf, et al., 2018 ). Additionally, vegetation typically adjusts biomass allocation by directing more resources to the underground parts to capture additional water in deep soil layers during drought period, thereby reducing water stress, with the woody component receiving a minimum biomass allocation (Eziz et al., 2017 ). Moreover, the growth of vegetation is more influenced by functional processes related to building a carbon sink (such as the growth of tissues like cambium) than by the quantity of biomass synthesized through photosynthesis (Sarris et al., 2007 ). Meanwhile, the vegetation leaves, serving as the site for photosynthesis in plants, are highly sensitive to the photosynthetic process. Consequently, the woody component is most sensitive to drought compared to litterfall and below-ground carbon (Brando et al., 2008 ). This high sensitivity of the woody component to drought and the functional processes related to its growth lead to more reduction in vegetation growth during drought compared to that of vegetation leaves (Gazol et al., 2018 ). Therefore, the recovery of the woody component is slower compared to the leaf component (Fig. 2 ). The moisture-related climatic conditions during the recovery period have a substantial impact on the recovery of the vegetation woody component. The mean monthly VPD, precipitation, and the soil moisture of layer 2 and layer 3 were identified as the most influencing climatic factors affecting the recovery of the vegetation woody component (Fig. 3 ). The strong impact of moisture-related climatic conditions during the recovery period on the recovery of woody component partly depends on the sensitivity of vegetation to specific climatic factors. For example, VPD modulates hydraulic function and structure in tropical rainforests during the recovery period with sufficient soil water supply (Binks et al., 2023 ), making tall Amazonian forests more sensitive to VPD than precipitation (Giardina et al., 2018 ). Soil moisture supplies the water resource to tropical forest and thus to be a key controller of tropical forest local hydrology (Bruno et al., 2006 ). The adaptability of vegetation to normal climate conditions also influences the recovery of vegetation woody components. Vegetation tends to become more adaptive to climate change in regions with higher seasonal variations (Walther et al., 2002 ). Thus, the EBF of South America experience greater seasonal variations with higher monthly precipitation and temperation variation (Fig. 4 g-h), allowing vegetation to adapt well to climate changes and correspondingly resulting in shorter recovery times (Fig. 2 – 3 ). Moreover, the EBF in Africa is more accustomed to drought and therefore exhibits greater resilience to droughts (Bennett et al., 2021 ), which also implies that when Africa experiences severe drought events, it will require more time for recovery (Fig. 2 ). There is a significant positive correlation between biomass and the recovery of the woody component in tropical EBF (Fig. 3 a). As biomass increases, the duration required for recovery also increases. This tendency may be attributed to higher biomass levels in the study area aligning with enhanced ecosystem diversity, facilitated by the diversity of plant hydraulic strategies and traits that can buffer a forest ecosystem against drought. Therefore, heightened biodiversity correlates with increased resilience to drought (Anderegg, et al., 2018 ), resulting in prolonged recovery periods. However, in this study, since biomass is not the most important influencing factor on the recovery of the woody component, it is therefore not appropriate to solely compare the recovery times of the woody vegetation component in different regions based on biomass alone. VOD data from different frequencies enables monitoring of vegetation recovery in tropical forests, but some limitations still exist. In tropical regions with dense forests, L-VOD may not fully represent changes in vegetation biomass (Dou et al., 2023 )d VOD and EVI also exhibit saturation problems in these areas (Du et al., 2017 ; Huete et al., 2002 ). Moreover, the smaller scattering effect during a drought period or dry season compared to normal conditions may result in elevated VOD estimations (Wang et al., 2023 ), thereby delaying the immediate observation of the drought's impact on vegetation. Also, since this study focuses on recovery time calculations at a monthly scale, we cannot well compare the recovery time of the upper canopy and leaves in nearly half of the EBF regions as they were observed to have fully recovered within one month. Despite these challenges, VOD data from different frequencies possess unique advantages in observing vegetation structures, and further improving the accuracy of VOD estimations is expected to bring additional benefits in research focusing on tropical rainforest areas, including but not limited to vegetation biomass carbon estimation and vegetation water content variation monitoring. 5 Conclusion In this study, we employed L-VOD, X-VOD, and EVI as proxies for the woody component, the upper canopy layer, and the leaf component of vegetation, respectively, to examine the recovery time of these vegetation elements in tropical forest regions following the 2015–2016 El Niño-induced drought. We found that the leaf component demonstrated a quicker recovery from drought, followed by the upper canopy layer and the woody component. Notably, the recovery time of the vegetation woody component exhibited greater spatial heterogeneity than the other two vegetation components. The recovery time of the woody component was primarily influenced by the impact of drought on the woody component, followed by the moisture-related climatic conditions during the recovery period (i.e., VPD, precipitation, and soil moisture) and the magnitude of the seasonal variation (i.e. the magnitude of the standard deviations in monthly temperature and precipitation). The more severe the damage to the woody component of vegetation during a drought, the less favorable the climate conditions during the recovery period (i.e., less precipitation, lower soil moisture, and higher VPD), and the higher the seasonal variations (i.e. larger standard deviations in monthly temperature and precipitation), the longer the corresponding recovery time for the woody component of vegetation. Therefore, due to the more extensive impact on the woody component during drought in Africa compared to South America, coupled with less favorable moisture-related climatic conditions during the post-El Niño recovery period and lower seasonal variation (i.e. larger standard deviations in monthly temperature and precipitation) in Africa than in South America, the recovery time of the woody component in Africa exceeded that in South America. Declarations Acknowledge This work is supported by the National Key Research and Development Program of China (Grant No. 2023YFF1303702) and the National Natural Science Foundation of China (Grant No. 42001299). References Anderegg, W. R. L., Konings, A. G., Trugman, A. T., Yu, K., Bowling, D. R., Gabbitas, R., Karp, D. S., Pacala, S., Sperry, J. S., Sulman, B. N., & Zenes, N. (2018). Hydraulic diversity of forests regulates ecosystem resilience during drought. Nature , 561 (7724), 538–541. https://doi.org/10.1038/s41586-018-0539-7 Anderegg, W. R. L., Wolf, A., Arango‐Velez, A., Choat, B., Chmura, D. J., Jansen, S., Kolb, T., Li, S., Meinzer, F. C., Pita, P., Resco De Dios, V., Sperry, J. S., Wolfe, B. T., & Pacala, S. (2018). Woody plants optimise stomatal behaviour relative to hydraulic risk. Ecology Letters , 21 (7), 968–977. https://doi.org/10.1111/ele.12962 Beck, H. E., Wood, E. F., Pan, M., Fisher, C. K., Miralles, D. G., van Dijk, A. I. J. M., McVicar, T. R., & Adler, R. F. (2019). MSWEP V2 Global 3-Hourly 0.1° Precipitation: Methodology and Quantitative Assessment. Bulletin of the American Meteorological Society , 100 (3), 473–500. https://doi.org/10.1175/BAMS-D-17-0138.1 Beer, C., Reichstein, M., Tomelleri, E., Ciais, P., Jung, M., Carvalhais, N., Rödenbeck, C., Arain, M. A., Baldocchi, D., Bonan, G. B., Bondeau, A., Cescatti, A., Lasslop, G., Lindroth, A., Lomas, M., Luyssaert, S., Margolis, H., Oleson, K. W., Roupsard, O., … Papale, D. (2010). Terrestrial Gross Carbon Dioxide Uptake: Global Distribution and Covariation with Climate. Science , 329 (5993), 834–838. https://doi.org/10.1126/science.1184984 Bennett, A. C., Dargie, G. C., Cuni-Sanchez, A., Tshibamba Mukendi, J., Hubau, W., Mukinzi, J. M., Phillips, O. L., Malhi, Y., Sullivan, M. J. P., Cooper, D. L. M., Adu-Bredu, S., Affum-Baffoe, K., Amani, C. A., Banin, L. F., Beeckman, H., Begne, S. K., Bocko, Y. E., Boeckx, P., Bogaert, J., … Lewis, S. L. (2021). Resistance of African tropical forests to an extreme climate anomaly. Proceedings of the National Academy of Sciences , 118 (21), e2003169118. https://doi.org/10.1073/pnas.2003169118 Binks, O., Cernusak, L. A., Liddell, M., Bradford, M., Coughlin, I., Bryant, C., Palma, A. C., Hoffmann, L., Alam, I., Carle, H. J., Rowland, L., Oliveira, R. S., Laurance, S. G. W., Mencuccini, M., & Meir, P. (2023). Vapour pressure deficit modulates hydraulic function and structure of tropical rainforests under nonlimiting soil water supply. New Phytologist , 240 (4), 1405–1420. https://doi.org/10.1111/nph.19257 Bonal, D., Burban, B., Stahl, C., Wagner, F., & Hérault, B. (2016). The response of tropical rainforests to drought—Lessons from recent research and future prospects. Annals of Forest Science , 73 (1), 27–44. https://doi.org/10.1007/s13595-015-0522-5 Brando, P. M., Nepstad, D. C., Davidson, E. A., Trumbore, S. E., Ray, D., & Camargo, P. (2008). Drought effects on litterfall, wood production and belowground carbon cycling in an Amazon forest: Results of a throughfall reduction experiment. Philosophical Transactions of the Royal Society B: Biological Sciences , 363 (1498), 1839–1848. https://doi.org/10.1098/rstb.2007.0031 Breiman, L. (2001). Random Forests. Machine Learning , 45 (1), 5–32. https://doi.org/10.1023/A:1010933404324 Bruno, R. D., Da Rocha, H. R., De Freitas, H. C., Goulden, M. L., & Miller, S. D. (2006). Soil moisture dynamics in an eastern Amazonian tropical forest. Hydrological Processes , 20 (12), 2477–2489. https://doi.org/10.1002/hyp.6211 Delgado-Baquerizo, M., Eldridge, D. J., Maestre, F. T., Karunaratne, S. B., Trivedi, P., Reich, P. B., & Singh, B. K. (2017). Climate legacies drive global soil carbon stocks in terrestrial ecosystems. Science Advances , 3 (4), e1602008. https://doi.org/10.1126/sciadv.1602008 Dou, Y., Tian, F., Wigneron, J.-P., Tagesson, T., Du, J., Brandt, M., Liu, Y., Zou, L., Kimball, J. S., & Fensholt, R. (2023). Reliability of using vegetation optical depth for estimating decadal and interannual carbon dynamics. Remote Sensing of Environment , 285 , 113390. https://doi.org/10.1016/j.rse.2022.113390 Du, J., Kimball, J. S., Jones, L. A., Kim, Y., Glassy, J., & Watts, J. D. (2017). A global satellite environmental data record derived from AMSR-E and AMSR2 microwave Earth observations. Earth System Science Data , 9 (2), 791–808. https://doi.org/10.5194/essd-9-791-2017 Eziz, A., Yan, Z., Tian, D., Han, W., Tang, Z., & Fang, J. (2017). Drought effect on plant biomass allocation: A meta‐analysis. Ecology and Evolution , 7 (24), 11002–11010. https://doi.org/10.1002/ece3.3630 Frappart, F., Wigneron, J.-P., Li, X., Liu, X., Al-Yaari, A., Fan, L., Wang, M., Moisy, C., Le Masson, E., Aoulad Lafkih, Z., Vallé, C., Ygorra, B., & Baghdadi, N. (2020). Global Monitoring of the Vegetation Dynamics from the Vegetation Optical Depth (VOD): A Review. Remote Sensing , 12 (18), 2915. https://doi.org/10.3390/rs12182915 Gazol, A., Camarero, J. J., Vicente‐Serrano, S. M., Sánchez‐Salguero, R., Gutiérrez, E., De Luis, M., Sangüesa‐Barreda, G., Novak, K., Rozas, V., Tíscar, P. A., Linares, J. C., Martín‐Hernández, N., Martínez Del Castillo, E., Ribas, M., García‐González, I., Silla, F., Camisón, A., Génova, M., Olano, J. M., … Galván, J. D. (2018). Forest resilience to drought varies across biomes. Global Change Biology , 24 (5), 2143–2158. https://doi.org/10.1111/gcb.14082 Gebrechorkos, S. H., Peng, J., Dyer, E., Miralles, D. G., Vicente-Serrano, S. M., Funk, C., Beck, H. E., Asfaw, D. T., Singer, M. B., & Dadson, S. J. (2023). Global high-resolution drought indices for 1981–2022. Earth System Science Data , 15 (12), 5449–5466. https://doi.org/10.5194/essd-15-5449-2023 Gerard, F. F., George, C. T., Hayman, G., Chavana‐Bryant, C., & Weedon, G. P. (2020). Leaf phenology amplitude derived from MODIS NDVI and EVI: Maps of leaf phenology synchrony for Meso‐ and South America. Geoscience Data Journal , 7 (1), 13–26. https://doi.org/10.1002/gdj3.87 Giardina, F., Konings, A. G., Kennedy, D., Alemohammad, S. H., Oliveira, R. S., Uriarte, M., & Gentine, P. (2018). Tall Amazonian forests are less sensitive to precipitation variability. Nature Geoscience , 11 (6), 405–409. https://doi.org/10.1038/s41561-018-0133-5 Hahn, C., Lüscher, A., Ernst-Hasler, S., Suter, M., & Kahmen, A. (2021). Timing of drought in the growing season and strong legacy effects determine the annual productivity of temperate grasses in a changing climate. Biogeosciences , 18 (2), 585–604. https://doi.org/10.5194/bg-18-585-2021 Hansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S. A., Tyukavina, A., Thau, D., Stehman, S. V., Goetz, S. J., Loveland, T. R., Kommareddy, A., Egorov, A., Chini, L., Justice, C. O., & Townshend, J. R. G. (2013). High-Resolution Global Maps of 21st-Century Forest Cover Change. Science , 342 (6160), 850–853. https://doi.org/10.1126/science.1244693 He, B., Xie, X., & Guo, L. (2023). A Shift From Temperature to Water as the Primary Driver for Interannual Variability of the Tropical Carbon Cycle. Geophysical Research Letters , 50 (15), e2023GL102812. https://doi.org/10.1029/2023GL102812 Huete, A., Didan, K., Miura, T., Rodriguez, E. P., Gao, X., & Ferreira, L. G. (2002). Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sensing of Environment , 83 (1–2), 195–213. https://doi.org/10.1016/S0034-4257(02)00096-2 Li, X., Wigneron, J.-P., Frappart, F., Fan, L., Ciais, P., Fensholt, R., Entekhabi, D., Brandt, M., Konings, A. G., Liu, X., Wang, M., Al-Yaari, A., & Moisy, C. (2021). Global-scale assessment and inter-comparison of recently developed/reprocessed microwave satellite vegetation optical depth products. Remote Sensing of Environment , 253 , 112208. https://doi.org/10.1016/j.rse.2020.112208 Li, X., Wigneron, J.-P., Frappart, F., Lannoy, G. D., Fan, L., Zhao, T., Gao, L., Tao, S., Ma, H., Peng, Z., Liu, X., Wang, H., Wang, M., Moisy, C., & Ciais, P. (2022). The first global soil moisture and vegetation optical depth product retrieved from fused SMOS and SMAP L-band observations. Remote Sensing of Environment , 282 , 113272. https://doi.org/10.1016/j.rse.2022.113272 Liang, J., Crowther, T. W., Picard, N., Wiser, S., Zhou, M., Alberti, G., Schulze, E.-D., McGuire, A. D., Bozzato, F., Pretzsch, H., de-Miguel, S., Paquette, A., Hérault, B., Scherer-Lorenzen, M., Barrett, C. B., Glick, H. B., Hengeveld, G. M., Nabuurs, G.-J., Pfautsch, S., … Reich, P. B. (2016). Positive biodiversity-productivity relationship predominant in global forests. Science , 354 (6309), aaf8957. https://doi.org/10.1126/science.aaf8957 Liu, J., Bowman, K. W., Schimel, D. S., Parazoo, N. C., Jiang, Z., Lee, M., Bloom, A. A., Wunch, D., Frankenberg, C., Sun, Y., O’Dell, C. W., Gurney, K. R., Menemenlis, D., Gierach, M., Crisp, D., & Eldering, A. (2017). Contrasting carbon cycle responses of the tropical continents to the 2015–2016 El Niño. Science , 358 (6360), eaam5690. https://doi.org/10.1126/science.aam5690 Liu, L., Chen, X., Ciais, P., Yuan, W., Maignan, F., Wu, J., Piao, S., Wang, Y., Wigneron, J., Fan, L., Gentine, P., Yang, X., Gong, F., Liu, H., Wang, C., Tang, X., Yang, H., Ye, Q., He, B., … Su, Y. (2022). Tropical tall forests are more sensitive and vulnerable to drought than short forests. Global Change Biology , 28 (4), 1583–1595. https://doi.org/10.1111/gcb.16017 Liu, Z., Zhu, J., Xia, J., & Huang, K. (2023). Declining resistance of vegetation productivity to droughts across global biomes. Agricultural and Forest Meteorology , 340 , 109602. https://doi.org/10.1016/j.agrformet.2023.109602 Loewenstein, N. J., & Pallardy, S. G. (1998). Drought tolerance, xylem sap abscisic acid and stomatal conductance during soil drying: A comparison of young plants of four temperate deciduous angiosperms. Tree Physiology , 18 (7), 421–430. https://doi.org/10.1093/treephys/18.7.421 Marengo, J. A., Tomasella, J., Alves, L. M., Soares, W. R., & Rodriguez, D. A. (2011). The drought of 2010 in the context of historical droughts in the Amazon region: DROUGHT AMAZON 2010. Geophysical Research Letters , 38 (12), n/a-n/a. https://doi.org/10.1029/2011GL047436 Sarris, D., Christodoulakis, D., & Körner, C. (2007). Recent decline in precipitation and tree growth in the eastern Mediterranean. Global Change Biology , 13 (6), 1187–1200. https://doi.org/10.1111/j.1365-2486.2007.01348.x Schwalm, C. R., Anderegg, W. R. L., Michalak, A. M., Fisher, J. B., Biondi, F., Koch, G., Litvak, M., Ogle, K., Shaw, J. D., Wolf, A., Huntzinger, D. N., Schaefer, K., Cook, R., Wei, Y., Fang, Y., Hayes, D., Huang, M., Jain, A., & Tian, H. (2017). Global patterns of drought recovery. Nature , 548 (7666), 202–205. https://doi.org/10.1038/nature23021 Taeger, S., Zang, C., Liesebach, M., Schneck, V., & Menzel, A. (2013). Impact of climate and drought events on the growth of Scots pine (Pinus sylvestris L.) provenances. Forest Ecology and Management , 307 , 30–42. https://doi.org/10.1016/j.foreco.2013.06.053 Tian, F., Brandt, M., Liu, Y. Y., Rasmussen, K., & Fensholt, R. (2017). Mapping gains and losses in woody vegetation across global tropical drylands. Global Change Biology , 23 (4), 1748–1760. https://doi.org/10.1111/gcb.13464 Vicente-Serrano, S. M., Beguería, S., & López-Moreno, J. I. (2010). A Multiscalar Drought Index Sensitive to Global Warming: The Standardized Precipitation Evapotranspiration Index. Journal of Climate , 23 (7), 1696–1718. https://doi.org/10.1175/2009JCLI2909.1 Vicente-Serrano, S. M., Gouveia, C., Camarero, J. J., Beguería, S., Trigo, R., López-Moreno, J. I., Azorín-Molina, C., Pasho, E., Lorenzo-Lacruz, J., Revuelto, J., Morán-Tejeda, E., & Sanchez-Lorenzo, A. (2013). Response of vegetation to drought time-scales across global land biomes. Proceedings of the National Academy of Sciences , 110 (1), 52–57. https://doi.org/10.1073/pnas.1207068110 Walther, G.-R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T. J. C., Fromentin, J.-M., Hoegh-Guldberg, O., & Bairlein, F. (2002). Ecological responses to recent climate change. Nature , 416 (6879), 389–395. https://doi.org/10.1038/416389a Wang, H., Wigneron, J.-P., Ciais, P., Yao, Y., Fan, L., Liu, X., Li, X., Green, J. K., Tian, F., Tao, S., Li, W., Frappart, F., Albergel, C., Wang, M., & Li, S. (2023). Seasonal variations in vegetation water content retrieved from microwave remote sensing over Amazon intact forests. Remote Sensing of Environment , 285 , 113409. https://doi.org/10.1016/j.rse.2022.113409 Wigneron, J.-P., Ciais, P., Li, X., Josep G, C., Tian, F., Wang, H., Bastos, A., Gatica, G., Kashyap, R., Liu, X., Sitch, S., Tao, S., Xiao, X., Carlo Espinoza, J., Frappart, F., Li, W., Qin, Y., Truchis De, T., & Fensholt, R. (2024). Global carbon balance of the forest: Satellite-based L-VOD results over the last decade and perspectives. Frontiers in Remote Sensing , 5 . https://doi.org/10.3389/frsen.2024.1338618 Wigneron, J.-P., Fan, L., Ciais, P., Bastos, A., Brandt, M., Chave, J., Saatchi, S., Baccini, A., & Fensholt, R. (2020). Tropical forests did not recover from the strong 2015–2016 El Niño event. Science Advances , 6 (6), eaay4603. https://doi.org/10.1126/sciadv.aay4603 Yao, Y., Liu, Y., Zhou, S., Song, J., & Fu, B. (2023). Soil moisture determines the recovery time of ecosystems from drought. Global Change Biology , 29 (13), 3562–3574. https://doi.org/10.1111/gcb.16620 Yuan, W., Zheng, Y., Piao, S., Ciais, P., Lombardozzi, D., Wang, Y., Ryu, Y., Chen, G., Dong, W., Hu, Z., Jain, A. K., Jiang, C., Kato, E., Li, S., Lienert, S., Liu, S., Nabel, J. E. M. S., Qin, Z., Quine, T., … Yang, S. (2019). Increased atmospheric vapor pressure deficit reduces global vegetation growth. Science Advances , 5 (8), eaax1396. https://doi.org/10.1126/sciadv.aax1396 Zhang, Z., Babst, F., Bellassen, V., Frank, D., Launois, T., Tan, K., Ciais, P., & Poulter, B. (2018). Converging Climate Sensitivities of European Forests Between Observed Radial Tree Growth and Vegetation Models. Ecosystems , 21 (3), 410–425. https://doi.org/10.1007/s10021-017-0157-5 Additional Declarations There is NO Competing Interest. Supplementary Files suplementary.docx Cite Share Download PDF Status: Published Journal Publication published 20 Nov, 2024 Read the published version in Communications Earth & Environment → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4464016","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":308275687,"identity":"444f7d48-873d-4947-809f-e378d9b9de34","order_by":0,"name":"Feng Tian","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4UlEQVRIiWNgGAWjYDACZhBhA8TsDWwgJmMDIR08YC1pINYBYrUwwLRIJBCpxZ6d+dkDhgSbPPnIN2aPeRhsZDccAIngdRibuQFDQlqx4e0cc2MehjTjDQdAIvj9YibB+ONw4sbZudukeRgOJ244wMMmgV8L+zcJhoT/iRtnngVp+U+MFh4zoJYDifMleEFaDhCh5TBPmURCQnLiBp78b5JzDJKNZx5mM8Orhb3/+DaJDwl2ifPbj6VJvKmwk+073vwMrxYwSABigwMgFiiomAmqhwL5BmJVjoJRMApGwYgDAAeQQC51AnAwAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-9686-2769","institution":"Wuhan University","correspondingAuthor":true,"prefix":"","firstName":"Feng","middleName":"","lastName":"Tian","suffix":""},{"id":308275688,"identity":"24503364-0fe8-4416-b56c-9e76f3029981","order_by":1,"name":"Yujie Dou","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yujie","middleName":"","lastName":"Dou","suffix":""},{"id":308275689,"identity":"8bdd9b87-1245-47ae-bd14-4b8e92b46c5f","order_by":2,"name":"Jean-Pierre Wigneron","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Jean-Pierre","middleName":"","lastName":"Wigneron","suffix":""},{"id":308275690,"identity":"c21429bf-187d-464d-9ded-5ca09c619f2a","order_by":3,"name":"Xiaojun Li","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Xiaojun","middleName":"","lastName":"Li","suffix":""},{"id":308275691,"identity":"a8de102a-11eb-45e9-b0d1-0d8f2f496b37","order_by":4,"name":"Wenmin Zhang","email":"","orcid":"https://orcid.org/0000-0001-6520-9559","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Wenmin","middleName":"","lastName":"Zhang","suffix":""},{"id":308275692,"identity":"8dc5c3e6-5c97-4963-8228-f06e11b81a19","order_by":5,"name":"Yaoliang Chen","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yaoliang","middleName":"","lastName":"Chen","suffix":""},{"id":308275693,"identity":"c4406082-6905-4ab2-8217-0ebb0e62dde3","order_by":6,"name":"Luwei Feng","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Luwei","middleName":"","lastName":"Feng","suffix":""},{"id":308275694,"identity":"f0817654-945a-4e5d-b555-0a72be405780","order_by":7,"name":"Qi Xie","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Qi","middleName":"","lastName":"Xie","suffix":""},{"id":308275695,"identity":"630ecc13-e53c-4052-9c3a-4bf66a135efd","order_by":8,"name":"Rasmus Fensholt","email":"","orcid":"https://orcid.org/0000-0003-3067-4527","institution":"University of Copenhagen","correspondingAuthor":false,"prefix":"","firstName":"Rasmus","middleName":"","lastName":"Fensholt","suffix":""}],"badges":[],"createdAt":"2024-05-23 03:55:06","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4464016/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4464016/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s43247-024-01892-9","type":"published","date":"2024-11-20T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":58111517,"identity":"3d73a21c-f874-4479-9889-8d0436c93bae","added_by":"auto","created_at":"2024-06-11 09:36:22","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":442610,"visible":true,"origin":"","legend":"\u003cp\u003eSee image above for figure legend\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-4464016/v1/8878c287175081d7f55e2fd7.png"},{"id":58111097,"identity":"c661ff81-9340-4a2c-9343-d2a2a2c90588","added_by":"auto","created_at":"2024-06-11 09:28:22","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":218823,"visible":true,"origin":"","legend":"\u003cp\u003eSpatial distribution of the recovery time from drought for the (a) woody component, (b) upper canopy layer, and (c) leaf component.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-4464016/v1/3e22f9f98fc204252870564a.png"},{"id":58111096,"identity":"88c39871-b5ba-4f96-8058-bdb7167c25b4","added_by":"auto","created_at":"2024-06-11 09:28:22","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":289797,"visible":true,"origin":"","legend":"\u003cp\u003eAttributions of the normal climatic conditions, climatic conditions during the woody recovery period, drought-related, and ecosystem-related factors to the recovery time of the woody component across tropical evergreen broadleaf forests. (a) The relative importance of the predictor variables in the random forest model is shown by the percentage increase of mean squared error (%IncMSE). The scatterplots illustrate the relationships between the woody recovery time and various factors, (b) L-VOD anomaly during the drought period, (c-d) mean monthly VPD and precipitation during the recovery period, (e) annual L-VOD pre-El Niño period, (f-g) monthly temperature and precipitation variation, (h-i) mean monthly soil moisture layer 2 and layer 3 during the recovery period, and (j) the long-term mean monthly precipitation.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-4464016/v1/601cfb6be4d07e18c8290445.png"},{"id":58111099,"identity":"014602c8-96f2-44c6-ba6c-41c5716c520f","added_by":"auto","created_at":"2024-06-11 09:28:22","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":458930,"visible":true,"origin":"","legend":"\u003cp\u003eStatistical distribution of the main influencing factors on the recovery of the woody component in South America and Africa, including (a) L-VOD anomaly during the drought period, (b-e) mean monthly VPD, precipitation, soil moisture of layer 2 and layer 3 during the recovery period, (f) annual L-VOD before El Niño, (g-h) seasonal variation in monthly temperature and precipitation variation, and (i) the long-term monthly mean precipitation. The horizontal lines at the top and bottom of the box plot represent the 25th and 75th percentiles, respectively, while the red line indicates the median.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-4464016/v1/484b21376c6c8e6fba0b982d.png"},{"id":69516334,"identity":"03f4a109-e95c-4314-92c0-e0a6ccad1acd","added_by":"auto","created_at":"2024-11-21 08:09:10","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1689017,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4464016/v1/a771254c-8023-47bc-b3ac-a08902dcdb57.pdf"},{"id":58111100,"identity":"80ee10aa-ca9f-4dcf-8e2c-ba0e764eb2b4","added_by":"auto","created_at":"2024-06-11 09:28:22","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":982141,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"suplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-4464016/v1/460b2ca3f8b2c44f65530136.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Woody component of tropical rainforest recovers slower from drought than the upper canopy layer and leaves","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eTropical ecosystems represent 34% of the global gross primary terrestrial productivity (Beer et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) and play a major role in carbon cycles at the global scale (Bonal et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). However, the effectiveness of capturing and storing carbon to mitigate future global warming partly depends on the impact of severe drought episodes as water is the primary determinant of the amount and allocation of forest biomass production, and thereby the interannual variability of the tropical carbon cycle (He et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Droughts in tropical regions are predominantly associated with the El Ni\u0026ntilde;o-Southern Oscillation (El Ni\u0026ntilde;o), and many extreme drought events in tropical regions coincide with El Ni\u0026ntilde;o events (Marengo et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). Notably the 2015\u0026ndash;2016 El Ni\u0026ntilde;o led to historically high temperatures and low precipitation across the tropics, and the growth rate of atmospheric carbon dioxide was the largest on record (Liu et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). An earlier study found that the carbon stocks in African and American humid forests had not recovered to pre- El Ni\u0026ntilde;o levels by 2017 (Wigneron et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), and the duration of the vegetation recovery period has yet to be determined.\u003c/p\u003e \u003cp\u003eDifferent vegetation components are characterized by differences in response time during drought conditions. Several experiments have demonstrated that the sensitivity of woody growth rate to drought surpasses that of vegetation canopy greenness (Gazol et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) because vegetation growth reduction is more mediated by the functional processes related to building a carbon sink than by the quantity of biomass synthesized through photosynthesis (Sarris et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Zhang et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). However, these experiments were conducted at the species level, and the spatial variability in the sensitivity of different parts of woody plants remains unexplored.\u003c/p\u003e \u003cp\u003eThe response of forests to drought does not only depend on forest resistance and adaptation strategies, but is also highly dependent on the severity of drought events (Taeger et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), the time scale at which drought occurs (Hahn et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), and the duration of the drought. For example, larger resistance to drought has been observed in spring when vegetation is in its reproductive stage, and productivity is at its peak (Hahn et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Tropical tall forests are found to be more sensitive and vulnerable to drought than short forests ( Liu et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and a higher species diversity could enhance drought resistance (Anderegg, Konings, et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Therefore, to better understand drought impacts on tropical forest ecosystems, it is necessary to consider and incorporate information on forest traits, drought severity, and climate to investigate the potential drivers of forest recovery from drought.\u003c/p\u003e \u003cp\u003eSince microwave observations at different frequencies can penetrate the vegetation layer at different depths, the microwave vegetation optical depth (VOD) data derived from multi-frequency microwave spaceborne observations provide a new way to investigate the sensitivity of vegetation to drought across various tree components. Lower frequencies (i.e. longer wavelengths) have deeper penetration depth because low-frequency radiation is less extinguished within the canopy. The L-band VOD (L-VOD) (Wigneron et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) has been demonstrated to convey essential information regarding the larger branches and trunks of woody components (Tian et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), while X-band VOD (X-VOD) is indicative of small branches and leaves of the upper canopy layer (Frappart et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Li et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Additionally, the Enhanced Vegetation Index (EVI) derived from optical satellite data is sensitive to vegetation greenness and can be considered a representation of leaf photosynthetic activity (Gerard et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Consequently, the information captured by L-VOD, X-VOD, and EVI, respectively, enables an exploration of the sensitivity of distinct vertical structures of the vegetation to drought conditions.\u003c/p\u003e \u003cp\u003eIn this study, we used satellite remote sensing data of L-VOD, X-VOD, and EVI as proxies for the woody component including branches and trunks, the upper layer of canopies, and the leaf component of tropical forest trees, respectively. To understand their respective sensitivity to drought, we aimed to investigate the recovery time of the different components of tropical evergreen forest (EBF) across the pan-tropics following the El Ni\u0026ntilde;o-induced drought of 2015\u0026ndash;2016, by analyzing time series of satellite data from 2010\u0026ndash;2022. We also employed the random forest method to ancillary data of climatic conditions, drought-related information, and ecosystem-related factors to investigate the primary drivers of spatial variability in the recovery time of the woody component in tropical EBF.\u003c/p\u003e"},{"header":"2 Data and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Vegetation indices data\u003c/h2\u003e \u003cp\u003eThe SMOS-SMAP-INRAE-BORDEAUX (SMOSMAP-IB) L-VOD product was retrieved from temperature brightness observed from ESA\u0026rsquo;s Soil Moisture Ocean Salinity (SMOS) and NASA\u0026rsquo;s Soil Moisture Active Passive (SMAP). It offers L-VOD at a semi-daily temporal resolution and a grid resolution of 25 km (Li et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The semi-daily global LPDR X-VOD dataset at 0.25\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(^\\circ\\)\u003c/span\u003e\u003c/span\u003e spatial resolution, derived from AMSR-E (Advanced Microwave Scanning Radiometer \u0026ndash; Earth Observing System) and AMSR-2 (Advanced Microwave Scanning Radiometer \u0026minus;\u0026thinsp;2) sensors (Du et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) was also employed. Both nighttime L-VOD (ascending orbit, 6:00 AM) and X-VOD (descending orbit, 1:30 AM) were aggregated to monthly data by averaging, covering the period from 2010 to 2022. The MODIS EVI data Version 6.1 from 2010 to 2022 was used, with monthly EVI data aggregated by taking the maximum values. All the vegetation index data have been converted to a geographic coordinate system format (i.e., WGS1984), with a spatial resolution of 0.25\u0026deg;.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 Drought data\u003c/h2\u003e \u003cp\u003eWe used the Standardized Precipitation Evapotranspiration Index (SPEI) as a drought indicator to assess the emergence, length, and intensity of drought events. As the humid forest has been shown to respond to drought within three months (Vicente-Serrano et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), the SPEI03 data with 0.05\u0026deg; spatial resolution (Gebrechorkos et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) was selected, which was calculated from Multi-Source Weighted-Ensemble Precipitation (MSWEP) and potential evapotranspiration (PET) from the Global Land Evaporation Amsterdam Model (GLEAM). The SPEI03 is calculated by factoring in the past 3-month aggregated precipitation and potential evapotranspiration, thus reflecting relatively short-term moisture conditions (Vicente-Serrano et al., \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). The SPEI data were resampled to the spatial resolution of VOD data at 0.25\u0026deg; by averaging.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e2.3 Ancillary data\u003c/h2\u003e \u003cp\u003eMonthly air temperature, dewpoint temperature, and soil moisture data at 0.1\u0026deg; resolution from 2010 to 2022 were taken from the ERA5 monthly average reanalysis dataset. The soil moisture data covers three layers, including layer 1 (0\u0026ndash;7 cm), layer 2 (7\u0026ndash;28 cm), and layer 3 (28\u0026ndash;100 cm). The air temperature and dewpoint temperature were used to calculate the vapor pressure deficit (VPD) using the method provided by Yuan et al., (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The precipitation data was derived from the MSWEP product with a 3 hours temporal resolution and 0.1\u0026deg; spatial resolution from 1979 to the present (Beck et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). All the climate data mentioned above were resampled to the spatial resolution of VOD data by averaging.\u003c/p\u003e \u003cp\u003eWe included the \u0026ldquo;elasticity of substitution\u0026rdquo; data, which reflects the degree to which various species can substitute each other in enhancing forest productivity (Liang et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), the magnitude of the intrinsic variability of vegetation water content data to represent vegetation water buffering (Dou et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), and a proxy for vegetation biomass data (i.e. the mean annual L-VOD pre- El Ni\u0026ntilde;o year). Note that the mean annual VOD before El Ni\u0026ntilde;o was computed as the 95th percentile of nighttime VOD from 2010 to 2014 as recommended by Dou et al., (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). MODIS landcover Version 6.1 data were used to delineate the EBF regions. Areas within VOD footprints where urban and cropland cover exceeds 5% were masked. Additionally, considering the potential impact of urban and cropland expansion on vegetation, we also excluded pixels with proportions of urban and cropland expansion exceeding 5% and pixels with forest loss exceeding 5% based on Hansen\u0026rsquo;s global forest change data (Hansen et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e2.4 Calculation of the recovery time\u003c/h2\u003e \u003cp\u003eMonthly SPEI and vegetation indices (VIs, i.e. L-VOD, X-VOD, and EVI) were used together to identify drought events and the calculation of recovery time for different vegetation components at pixel-scale (Schwalm et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Yao et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). The monthly VIs time-series data were deseasonalized by subtracting the monthly average values (calculated from the full period excluding the drought years from 2015 to 2016) from the VIs time series to remove the effects of the seasonal cycle and then detrended to eliminate the long-term trend. When a drought event happens and the standardized deviation (SD) of the detrended VI data falls below \u0026minus;\u0026thinsp;0.5 SD, the vegetation is considered to have been negatively affected by the drought event (Yao et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe drought event was considered to begin when the SPEI was lower than \u0026minus;\u0026thinsp;1 and the detrended VI data at the same time were below \u0026minus;\u0026thinsp;0.5 SD, and it ended when the SPEI was higher than \u0026minus;\u0026thinsp;1 and the detrended VIs data remained below \u0026minus;\u0026thinsp;0.5 SD, or SPEI stayed lower than \u0026minus;\u0026thinsp;1 but the detrended VIs data increased above \u0026minus;\u0026thinsp;0.5 SD. Additionally, we only focused on drought events lasting for at least 2 months.\u003c/p\u003e \u003cp\u003eThe calculation of the drought recovery time for different vegetation components is based on the following criteria (Yao et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e): (Yao et al., \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e): (1) If the detrended VI data reached a local minimum during the drought period (see above), the recovery time was defined as the period from the time when the detrended VI data reached the minimum value to the time when the detrended VI data were higher than \u0026minus;\u0026thinsp;0.5 SD. (2) If the condition above was not met, the recovery time was defined as the period from the end of the drought event (see above) to the time when the detrended VI data were larger than \u0026minus;\u0026thinsp;0.5 SD.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e2.5 Drivers of the recovery of vegetation woody component\u003c/h2\u003e \u003cp\u003eThe leaf component generally represents only a small fraction of the entire above-ground woody biomass and may not be representative of the trends and spatial patterns of the woody component (Tian et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Thus, only the drivers of the recovery of vegetation woody components were investigated.\u003c/p\u003e \u003cp\u003eWe included 44 response variables in the random forest regression model to investigate the relative significance of these variables to the recovery time of the woody component. The 44 response variables were reclassified into four classes, including the normal climatic conditions, the climatic conditions during the recovery period, drought-related factors, and ecosystem-related factors.\u003c/p\u003e \u003cp\u003eSpecifically, the variables covering the normal climatic conditions comprise annual mean, 25th percentile minimum, 75th percentile maximum, and standard deviation of air temperatures (T_mean, T_min25, T_max75, T_std), precipitation (P_mean, P_min25, P_max75, P_std), soil moisture from layers 1 to 3 (SM1_mean, SM1_min25, SM1_max75, SM1_std, SM2_mean, SM2_min25, SM2_max75, SM2_std, SM3_mean, SM3_min25, SM3_max75, SM3_std), and VPD (VPD_mean, VPD_min25, VPD_max75, VPD_std), covering the period from 2010 to 2022, excluding 2015 to 2016. The standard deviation of air temperatures, precipitation, soil moisture, and VPD were defined as the seasonal variation of normal climate conditions in this study.\u003c/p\u003e \u003cp\u003eThe variables denoting the climatic conditions during the recovery period include monthly mean air temperature (Recovery_T_mean), precipitation (Recovery_P_mean), VPD (Recovery_VPD_mean), and soil moisture from layers 1 to 3 (Recovery_SM1_mean, Recovery_SM2_mean, Recovery_SM3_mean).\u003c/p\u003e \u003cp\u003eDrought-related factors encompass drought duration, drought severity (i.e. mean SPEI), the number of dry months (monthly precipitation less than 100 mm), anomalies of the climatic conditions relative to the pre-El Ni\u0026ntilde;o period, including temperature (T_anomaly), precipitation (P_anomaly), VPD (VPD_anomaly), and soil moisture from layer1 to layer3 (SM1_anomaly, SM2_anomaly, SM3_anomaly), and the severity of the drought impact on the woody component (i.e. the L-VOD anomaly during the drought period relative to the pre-drought condition, L-VOD_anomaly).\u003c/p\u003e \u003cp\u003eEcosystem-related variables include the magnitude of the intrinsic variability of vegetation water content data representing vegetation water buffering (Mean_delta_VOD_day), biomass (i.e. pre-El Ni\u0026ntilde;o annual VOD values, annual L-VOD), and the elasticity of substitution.\u003c/p\u003e \u003cp\u003eThe random forest regression model can explain interactions and nonlinear relationships between predictors (Breiman, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). The importance of each response variable was assessed through the percentage increase in the mean square error (%IncMSE) between target and response values (Delgado-Baquerizo et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The values of the %IncMSE were generated from a random forest model consisting of 500 decision trees in this study, and higher values of %IncMSE suggest higher importance of the response variables. It should be noted that random forest is a tree-based ensemble model that is sensitive to the nonlinear relationships and interactions among features. Therefore, even if a feature shows high importance in terms of \"%IncMSE\", its coefficient of determination between recovery time and the response variables may not necessarily be high (Liu et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Additionally, we conducted a principal component analysis (PCA) on the input data to transform highly correlated input factors (e.g., precipitation and VPD) into a set of uncorrelated principal components to mitigate the impact of collinearity. These principal components were used as the new input features for training the random forest model.\u003c/p\u003e \u003c/div\u003e"},{"header":"3 Results","content":"\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 The severity of drought\u003c/h2\u003e\n \u003cp\u003eMost EBF in South America (56%) and Africa (90%) regions have experienced severe drought (i.e., SPEI \u0026lt; -1.5) (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003ea) caused by the 2015\u0026ndash;2016 El Ni\u0026ntilde;o, and there were obvious differences in the drought duration across the pantropical area (Fig. \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eb). The EBF in Africa showed the most widespread exposure to long-duration droughts, with drought periods lasting up to 6 months covering 93% of the region, whereas 58% of South America forests and 40% of Asian forests have been exposed to such long-duration droughts, respectively.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Recovery time of different vegetation components\u003c/h2\u003e\n \u003cp\u003eNoticeable spatial differences in the recovery time of the woody component (branches and woody), the upper canopy layer, and the leaf component were observed following the 2015\u0026ndash;2016 drought (Figure. 2). The recovery of the upper canopy layer and leaves was faster than the woody component. Nearly 73% of the upper canopy layer area and 88% of the leaf area recovered to the pre- El Ni\u0026ntilde;o conditions within two months, while only 52% of the drought-affected woody component area showed a similar recovery time (Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003ea). Moreover, there were 29% of the area for the woody component did not recover within one year (Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003ea). with 73% of the area having recovered within two months for the upper canopy layer and 88% in the case of leaves.\u003c/p\u003e\n \u003cp\u003eThe recovery time of the upper canopy layer and the leaf component exhibited less spatial variation, but there were notable variations in the recovery time of the woody components in the forest regions of South America, Africa, and Asia. The woody components of EBF in South America show the fastest recovery (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003ea), with 57% of the region recovering within 2 months. The recovery time for EBF in Africa is the longest, with 42% of the region requiring more than 12 months to fully recover.\u003c/p\u003e\n \u003cp\u003eAs for the recovery time for different vegetation components, the pixel scale leaf component recovers first, followed by the upper canopy layer and the woody components (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e, Fig. \u003cspan class=\"InternalRef\"\u003eS1\u003c/span\u003e-S2). In 89% of forested areas, the leaf component showed simultaneous recovery time with either the upper canopy layer, the woody component, or both (in 12% of the study area, the leaf component recovered first preceding the recovery of the upper canopy layer and the woody component) (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eb). This phenomenon was common in South America, Africa, and Asia (Fig. S2). The regions where the upper canopy layer showed simultaneous recovery time with either the leaf component, the woody component, or both account for 74% (in 4% of forested areas, the upper canopy layer recovered first preceding the recovery of the other two components). Nearly 20% of forested areas showed simultaneous recovery time of the woody component with either the leaf component, the upper canopy layer, or both (the areas where the woody component recovered first accounts for 2%).\u003c/p\u003e\n \u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3 Main drivers of woody component recovery across global tropical rainforests\u003c/h2\u003e\n \u003cp\u003eThe recovery time of the upper canopy layer and the leaf component exhibited less spatial variation in recovery time (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e), so we solely investigated the factors influencing the recovery time of the woody component.\u003c/p\u003e\n \u003cp\u003eThe input factors incorporated in the random forest model collectively explained 70% of the variability in the woody component recovery across the tropical region of EBF forests. The severity of the drought impact on the woody component (i.e. L-VOD anomaly) emerged as the primary determinant of the recovery time (%IncMSE\u0026thinsp;=\u0026thinsp;102%). Moreover, there was a significant negative correlation between L-VOD anomalies and the recovery time of the woody component (R = -0.58, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01), indicating that more severe impacts of drought on the woody component result in longer recovery time (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eb). Subsequently, the climatic conditions related to moisture levels during the recovery period, including mean monthly VPD (%IncMSE\u0026thinsp;=\u0026thinsp;32%, R = -0.13, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and mean monthly precipitation (%IncMSE\u0026thinsp;=\u0026thinsp;29%, R\u0026thinsp;=\u0026thinsp;0.03, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05), were found to be important explanatory variables of the recovery time. Biomass (i.e. annual L-VOD) is also closely related to the recovery of vegetation woody components (%IncMSE\u0026thinsp;=\u0026thinsp;26%, R\u0026thinsp;=\u0026thinsp;0.27, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01). The higher the vegetation biomass, the longer the corresponding recovery time.\u003c/p\u003e\n \u003cp\u003eThe seasonal variation (i.e., the standard deviation of monthly air temperature and precipitation) also played an important role in influencing the recovery of the vegetation woody component. When the seasonal showed minor variations (e.g., smaller standard deviations in monthly temperature (%IncMSE\u0026thinsp;=\u0026thinsp;32%, R = -0.24, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) and monthly precipitation (%IncMSE\u0026thinsp;=\u0026thinsp;32%, R = -0.29, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01)), the recovery time of the woody layer tended to be prolonged. Finally, soil moisture of layer 2 during the recovery period (%IncMSE\u0026thinsp;=\u0026thinsp;20%, R\u0026thinsp;=\u0026thinsp;0.02, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05), soil moisture of layer 3 during the recovery period (%IncMSE\u0026thinsp;=\u0026thinsp;19%, R = -0.03, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05), and the long-term mean monthly precipitation (%IncMSE\u0026thinsp;=\u0026thinsp;19%, R = -0.12, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.01) also positively contributed to the recovery of the woody component, with more water availability corresponded to shorter recovery time.\u003c/p\u003e\n \u003cp\u003eWe also analyzed the primary influencing factors on the recovery time of EBF in South America (56% explanation), Africa (73% explanation), and Asia (54% explanation) separately (Fig. S3 - S5). We found that the severity of the drought impact on the woody component (i.e. L-VOD anomaly) consistently remained the most significant factor affecting the recovery time of the woody component over all three continents. This was followed by the climatic conditions during the recovery period in South America and Africa, as well as the drought-related factors (e.g. air temperature anomaly, VPD anomaly) in Asia.\u003c/p\u003e\n \u003cp\u003e\u003cspan\u003e\u003cstrong\u003e3.4 Possible explanations for the faster recovery of the woody component in South America compared to Africa\u003c/strong\u003e\u003cbr\u003e\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe primary drivers of the woody components recovery in rainforests worldwide could well explain why these components recovered more quickly in South America than in Africa. Firstly, the severity of the drought impact on the woody component (i.e. L-VOD anomaly) was more pronounced in Africa than in South America (mean L-VOD anomaly = -3.0 SD in Africa \u003cem\u003evs\u003c/em\u003e. -2.3 SD in South America) (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea). In Africa, the most negative impact of drought on the woody component was concentrated in the Congo Rainforest region, where the L-VOD anomaly was lower than in other regions (Fig. S6). Secondly, the climatic conditions related to moisture levels during the recovery period in South American EBF regions were more favorable than in Africa with lower VPD (mean VPD\u0026thinsp;=\u0026thinsp;5.9 in South America \u003cem\u003evs\u003c/em\u003e. 6.3 in Africa), more precipitation (mean precipitation\u0026thinsp;=\u0026thinsp;182 mm in South America \u003cem\u003evs\u003c/em\u003e. 136 mm in Africa), and more soil moisture (mean soil moisture of layer 2\u0026thinsp;=\u0026thinsp;0.39 m\u003csup\u003e3\u003c/sup\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e in South America \u003cem\u003evs\u003c/em\u003e. 0.37 m\u003csup\u003e3\u003c/sup\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e in Africa, mean soil moisture of layer 3\u0026thinsp;=\u0026thinsp;0.39 m\u003csup\u003e3\u003c/sup\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e in South America \u003cem\u003evs\u003c/em\u003e. 0.35 m\u003csup\u003e3\u003c/sup\u003e m\u003csup\u003e\u0026minus;\u0026thinsp;3\u003c/sup\u003e in Africa) (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003eb-e). The pre-El Ni\u0026ntilde;o annual VOD in EBF of South America was slightly higher than in Africa (mean pre-El Ni\u0026ntilde;o annual VOD\u0026thinsp;=\u0026thinsp;0.91 in South America \u003cem\u003evs\u003c/em\u003e. 0.87 in Africa) (Fig. \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003ef), but the impact of drought on the woody component and the influence of climate on the recovery period is greater than that of biomass alone (Fig. \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003ea). Therefore, solely comparing biomass values cannot adequately explain why the recovery time of the woody vegetation layer in South America is faster than that in Africa.\u003c/p\u003e\n \u003cp\u003eFinally, the EBF in Africa showed a more stable seasonal variation with lower monthly temperature variation (mean temperature variation\u0026thinsp;=\u0026thinsp;0.9 SD in South America \u003cem\u003evs\u003c/em\u003e. 0.88 SD in Africa) and monthly precipitation variation (mean precipitation variation\u0026thinsp;=\u0026thinsp;106 SD in South America \u003cem\u003evs\u003c/em\u003e. 71 SD in Africa). Moreover, the EBF has lower precipitation availability in Africa than so (mean precipitation\u0026thinsp;=\u0026thinsp;207 mm in South America vs. 137 mm in Africa). Therefore, the severity of drought impact on the woody component, the less favorable moisture-related climatic conditions during recovery, and the smaller seasonal variation in Africa compared to South America\u0026apos;s EBF contribute to a slower recovery of the woody component in Africa compared to South America.\u003c/p\u003e\n \u003cp\u003eIt should be noted that due to the relatively smaller area of EBF in Asia compared to South America and Africa, we chose only to focus our analysis on understanding the differences in the recovery time of woody components in vegetation between South America and Africa.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eIn this study, we used L-VOD, X-VOD, and EVI data separately as proxies of the woody component (branches and woody), the upper canopy layer, and the leaf component to investigate the recovery of these components in tropical EBF regions during the extreme drought period induced by the 2015\u0026ndash;2016 El Ni\u0026ntilde;o event. We found that the recovery time of the leaf component is the fastest, followed by the upper canopy layer and the woody component (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e-S2).\u003c/p\u003e \u003cp\u003eDuring a drought period, the leaf component initially responds by closing stomata, followed by a reaction in the woody parts (Loewenstein \u0026amp; Pallardy, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e1998\u003c/span\u003e). While the longstanding theory holds that stomata optimize fitness by maintaining constant marginal water use efficiency over a specified time frame, a recent evolutionary theory suggests an alternative perspective that stomata aim to maximize the carbon gain while minimizing carbon costs and the risk of hydraulic damage (Anderegg, Wolf, et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Additionally, vegetation typically adjusts biomass allocation by directing more resources to the underground parts to capture additional water in deep soil layers during drought period, thereby reducing water stress, with the woody component receiving a minimum biomass allocation (Eziz et al., \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Moreover, the growth of vegetation is more influenced by functional processes related to building a carbon sink (such as the growth of tissues like cambium) than by the quantity of biomass synthesized through photosynthesis (Sarris et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Meanwhile, the vegetation leaves, serving as the site for photosynthesis in plants, are highly sensitive to the photosynthetic process. Consequently, the woody component is most sensitive to drought compared to litterfall and below-ground carbon (Brando et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). This high sensitivity of the woody component to drought and the functional processes related to its growth lead to more reduction in vegetation growth during drought compared to that of vegetation leaves (Gazol et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Therefore, the recovery of the woody component is slower compared to the leaf component (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe moisture-related climatic conditions during the recovery period have a substantial impact on the recovery of the vegetation woody component. The mean monthly VPD, precipitation, and the soil moisture of layer 2 and layer 3 were identified as the most influencing climatic factors affecting the recovery of the vegetation woody component (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The strong impact of moisture-related climatic conditions during the recovery period on the recovery of woody component partly depends on the sensitivity of vegetation to specific climatic factors. For example, VPD modulates hydraulic function and structure in tropical rainforests during the recovery period with sufficient soil water supply (Binks et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), making tall Amazonian forests more sensitive to VPD than precipitation (Giardina et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Soil moisture supplies the water resource to tropical forest and thus to be a key controller of tropical forest local hydrology (Bruno et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe adaptability of vegetation to normal climate conditions also influences the recovery of vegetation woody components. Vegetation tends to become more adaptive to climate change in regions with higher seasonal variations (Walther et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Thus, the EBF of South America experience greater seasonal variations with higher monthly precipitation and temperation variation (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eg-h), allowing vegetation to adapt well to climate changes and correspondingly resulting in shorter recovery times (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Moreover, the EBF in Africa is more accustomed to drought and therefore exhibits greater resilience to droughts (Bennett et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e), which also implies that when Africa experiences severe drought events, it will require more time for recovery (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThere is a significant positive correlation between biomass and the recovery of the woody component in tropical EBF (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). As biomass increases, the duration required for recovery also increases. This tendency may be attributed to higher biomass levels in the study area aligning with enhanced ecosystem diversity, facilitated by the diversity of plant hydraulic strategies and traits that can buffer a forest ecosystem against drought. Therefore, heightened biodiversity correlates with increased resilience to drought (Anderegg, et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), resulting in prolonged recovery periods. However, in this study, since biomass is not the most important influencing factor on the recovery of the woody component, it is therefore not appropriate to solely compare the recovery times of the woody vegetation component in different regions based on biomass alone.\u003c/p\u003e \u003cp\u003eVOD data from different frequencies enables monitoring of vegetation recovery in tropical forests, but some limitations still exist. In tropical regions with dense forests, L-VOD may not fully represent changes in vegetation biomass (Dou et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)d VOD and EVI also exhibit saturation problems in these areas (Du et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Huete et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Moreover, the smaller scattering effect during a drought period or dry season compared to normal conditions may result in elevated VOD estimations (Wang et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), thereby delaying the immediate observation of the drought's impact on vegetation. Also, since this study focuses on recovery time calculations at a monthly scale, we cannot well compare the recovery time of the upper canopy and leaves in nearly half of the EBF regions as they were observed to have fully recovered within one month. Despite these challenges, VOD data from different frequencies possess unique advantages in observing vegetation structures, and further improving the accuracy of VOD estimations is expected to bring additional benefits in research focusing on tropical rainforest areas, including but not limited to vegetation biomass carbon estimation and vegetation water content variation monitoring.\u003c/p\u003e"},{"header":"5 Conclusion","content":"\u003cp\u003eIn this study, we employed L-VOD, X-VOD, and EVI as proxies for the woody component, the upper canopy layer, and the leaf component of vegetation, respectively, to examine the recovery time of these vegetation elements in tropical forest regions following the 2015–2016 El Niño-induced drought. We found that the leaf component demonstrated a quicker recovery from drought, followed by the upper canopy layer and the woody component. Notably, the recovery time of the vegetation woody component exhibited greater spatial heterogeneity than the other two vegetation components. The recovery time of the woody component was primarily influenced by the impact of drought on the woody component, followed by the moisture-related climatic conditions during the recovery period (i.e., VPD, precipitation, and soil moisture) and the magnitude of the seasonal variation (i.e. the magnitude of the standard deviations in monthly temperature and precipitation). The more severe the damage to the woody component of vegetation during a drought, the less favorable the climate conditions during the recovery period (i.e., less precipitation, lower soil moisture, and higher VPD), and the higher the seasonal variations (i.e. larger standard deviations in monthly temperature and precipitation), the longer the corresponding recovery time for the woody component of vegetation. Therefore, due to the more extensive impact on the woody component during drought in Africa compared to South America, coupled with less favorable moisture-related climatic conditions during the post-El Niño recovery period and lower seasonal variation (i.e. larger standard deviations in monthly temperature and precipitation) in Africa than in South America, the recovery time of the woody component in Africa exceeded that in South America.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003eAcknowledge\u003c/p\u003e\n\u003cp\u003eThis work is supported by the National Key Research and Development Program of China (Grant No. 2023YFF1303702) and the National Natural Science Foundation of China (Grant No. 42001299).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAnderegg, W. R. L., Konings, A. G., Trugman, A. T., Yu, K., Bowling, D. R., Gabbitas, R., Karp, D. S., Pacala, S., Sperry, J. S., Sulman, B. N., \u0026amp; Zenes, N. (2018). Hydraulic diversity of forests regulates ecosystem resilience during drought. \u003cem\u003eNature\u003c/em\u003e, \u003cem\u003e561\u003c/em\u003e(7724), 538\u0026ndash;541. https://doi.org/10.1038/s41586-018-0539-7\u003c/li\u003e\n\u003cli\u003eAnderegg, W. R. L., Wolf, A., Arango‐Velez, A., Choat, B., Chmura, D. J., Jansen, S., Kolb, T., Li, S., Meinzer, F. C., Pita, P., Resco De Dios, V., Sperry, J. S., Wolfe, B. T., \u0026amp; Pacala, S. (2018). Woody plants optimise stomatal behaviour relative to hydraulic risk. \u003cem\u003eEcology Letters\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(7), 968\u0026ndash;977. https://doi.org/10.1111/ele.12962\u003c/li\u003e\n\u003cli\u003eBeck, H. E., Wood, E. F., Pan, M., Fisher, C. K., Miralles, D. G., van Dijk, A. I. J. M., McVicar, T. R., \u0026amp; Adler, R. F. (2019). MSWEP V2 Global 3-Hourly 0.1\u0026deg; Precipitation: Methodology and Quantitative Assessment. \u003cem\u003eBulletin of the American Meteorological Society\u003c/em\u003e, \u003cem\u003e100\u003c/em\u003e(3), 473\u0026ndash;500. https://doi.org/10.1175/BAMS-D-17-0138.1\u003c/li\u003e\n\u003cli\u003eBeer, C., Reichstein, M., Tomelleri, E., Ciais, P., Jung, M., Carvalhais, N., R\u0026ouml;denbeck, C., Arain, M. A., Baldocchi, D., Bonan, G. B., Bondeau, A., Cescatti, A., Lasslop, G., Lindroth, A., Lomas, M., Luyssaert, S., Margolis, H., Oleson, K. W., Roupsard, O., \u0026hellip; Papale, D. (2010). Terrestrial Gross Carbon Dioxide Uptake: Global Distribution and Covariation with Climate. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e329\u003c/em\u003e(5993), 834\u0026ndash;838. https://doi.org/10.1126/science.1184984\u003c/li\u003e\n\u003cli\u003eBennett, A. C., Dargie, G. C., Cuni-Sanchez, A., Tshibamba Mukendi, J., Hubau, W., Mukinzi, J. M., Phillips, O. L., Malhi, Y., Sullivan, M. J. P., Cooper, D. L. M., Adu-Bredu, S., Affum-Baffoe, K., Amani, C. A., Banin, L. F., Beeckman, H., Begne, S. K., Bocko, Y. E., Boeckx, P., Bogaert, J., \u0026hellip; Lewis, S. L. (2021). Resistance of African tropical forests to an extreme climate anomaly. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e118\u003c/em\u003e(21), e2003169118. https://doi.org/10.1073/pnas.2003169118\u003c/li\u003e\n\u003cli\u003eBinks, O., Cernusak, L. A., Liddell, M., Bradford, M., Coughlin, I., Bryant, C., Palma, A. C., Hoffmann, L., Alam, I., Carle, H. J., Rowland, L., Oliveira, R. S., Laurance, S. G. W., Mencuccini, M., \u0026amp; Meir, P. (2023). Vapour pressure deficit modulates hydraulic function and structure of tropical rainforests under nonlimiting soil water supply. \u003cem\u003eNew Phytologist\u003c/em\u003e, \u003cem\u003e240\u003c/em\u003e(4), 1405\u0026ndash;1420. https://doi.org/10.1111/nph.19257\u003c/li\u003e\n\u003cli\u003eBonal, D., Burban, B., Stahl, C., Wagner, F., \u0026amp; H\u0026eacute;rault, B. (2016). The response of tropical rainforests to drought\u0026mdash;Lessons from recent research and future prospects. \u003cem\u003eAnnals of Forest Science\u003c/em\u003e, \u003cem\u003e73\u003c/em\u003e(1), 27\u0026ndash;44. https://doi.org/10.1007/s13595-015-0522-5\u003c/li\u003e\n\u003cli\u003eBrando, P. M., Nepstad, D. C., Davidson, E. A., Trumbore, S. E., Ray, D., \u0026amp; Camargo, P. (2008). Drought effects on litterfall, wood production and belowground carbon cycling in an Amazon forest: Results of a throughfall reduction experiment. \u003cem\u003ePhilosophical Transactions of the Royal Society B: Biological Sciences\u003c/em\u003e, \u003cem\u003e363\u003c/em\u003e(1498), 1839\u0026ndash;1848. https://doi.org/10.1098/rstb.2007.0031\u003c/li\u003e\n\u003cli\u003eBreiman, L. (2001). Random Forests. \u003cem\u003eMachine Learning\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(1), 5\u0026ndash;32. https://doi.org/10.1023/A:1010933404324\u003c/li\u003e\n\u003cli\u003eBruno, R. D., Da Rocha, H. R., De Freitas, H. C., Goulden, M. L., \u0026amp; Miller, S. D. (2006). Soil moisture dynamics in an eastern Amazonian tropical forest. \u003cem\u003eHydrological Processes\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(12), 2477\u0026ndash;2489. https://doi.org/10.1002/hyp.6211\u003c/li\u003e\n\u003cli\u003eDelgado-Baquerizo, M., Eldridge, D. J., Maestre, F. T., Karunaratne, S. B., Trivedi, P., Reich, P. B., \u0026amp; Singh, B. K. (2017). Climate legacies drive global soil carbon stocks in terrestrial ecosystems. \u003cem\u003eScience Advances\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e(4), e1602008. https://doi.org/10.1126/sciadv.1602008\u003c/li\u003e\n\u003cli\u003eDou, Y., Tian, F., Wigneron, J.-P., Tagesson, T., Du, J., Brandt, M., Liu, Y., Zou, L., Kimball, J. S., \u0026amp; Fensholt, R. (2023). Reliability of using vegetation optical depth for estimating decadal and interannual carbon dynamics. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e, \u003cem\u003e285\u003c/em\u003e, 113390. https://doi.org/10.1016/j.rse.2022.113390\u003c/li\u003e\n\u003cli\u003eDu, J., Kimball, J. S., Jones, L. A., Kim, Y., Glassy, J., \u0026amp; Watts, J. D. (2017). A global satellite environmental data record derived from AMSR-E and AMSR2 microwave Earth observations. \u003cem\u003eEarth System Science Data\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e(2), 791\u0026ndash;808. https://doi.org/10.5194/essd-9-791-2017\u003c/li\u003e\n\u003cli\u003eEziz, A., Yan, Z., Tian, D., Han, W., Tang, Z., \u0026amp; Fang, J. (2017). Drought effect on plant biomass allocation: A meta‐analysis. \u003cem\u003eEcology and Evolution\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(24), 11002\u0026ndash;11010. https://doi.org/10.1002/ece3.3630\u003c/li\u003e\n\u003cli\u003eFrappart, F., Wigneron, J.-P., Li, X., Liu, X., Al-Yaari, A., Fan, L., Wang, M., Moisy, C., Le Masson, E., Aoulad Lafkih, Z., Vall\u0026eacute;, C., Ygorra, B., \u0026amp; Baghdadi, N. (2020). Global Monitoring of the Vegetation Dynamics from the Vegetation Optical Depth (VOD): A Review. \u003cem\u003eRemote Sensing\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(18), 2915. https://doi.org/10.3390/rs12182915\u003c/li\u003e\n\u003cli\u003eGazol, A., Camarero, J. J., Vicente‐Serrano, S. M., S\u0026aacute;nchez‐Salguero, R., Guti\u0026eacute;rrez, E., De Luis, M., Sang\u0026uuml;esa‐Barreda, G., Novak, K., Rozas, V., T\u0026iacute;scar, P. A., Linares, J. C., Mart\u0026iacute;n‐Hern\u0026aacute;ndez, N., Mart\u0026iacute;nez Del Castillo, E., Ribas, M., Garc\u0026iacute;a‐Gonz\u0026aacute;lez, I., Silla, F., Camis\u0026oacute;n, A., G\u0026eacute;nova, M., Olano, J. M., \u0026hellip; Galv\u0026aacute;n, J. D. (2018). Forest resilience to drought varies across biomes. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e(5), 2143\u0026ndash;2158. https://doi.org/10.1111/gcb.14082\u003c/li\u003e\n\u003cli\u003eGebrechorkos, S. H., Peng, J., Dyer, E., Miralles, D. G., Vicente-Serrano, S. M., Funk, C., Beck, H. E., Asfaw, D. T., Singer, M. B., \u0026amp; Dadson, S. J. (2023). Global high-resolution drought indices for 1981\u0026ndash;2022. \u003cem\u003eEarth System Science Data\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(12), 5449\u0026ndash;5466. https://doi.org/10.5194/essd-15-5449-2023\u003c/li\u003e\n\u003cli\u003eGerard, F. F., George, C. T., Hayman, G., Chavana‐Bryant, C., \u0026amp; Weedon, G. P. (2020). Leaf phenology amplitude derived from MODIS NDVI and EVI: Maps of leaf phenology synchrony for Meso‐ and South America. \u003cem\u003eGeoscience Data Journal\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(1), 13\u0026ndash;26. https://doi.org/10.1002/gdj3.87\u003c/li\u003e\n\u003cli\u003eGiardina, F., Konings, A. G., Kennedy, D., Alemohammad, S. H., Oliveira, R. S., Uriarte, M., \u0026amp; Gentine, P. (2018). Tall Amazonian forests are less sensitive to precipitation variability. \u003cem\u003eNature Geoscience\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(6), 405\u0026ndash;409. https://doi.org/10.1038/s41561-018-0133-5\u003c/li\u003e\n\u003cli\u003eHahn, C., L\u0026uuml;scher, A., Ernst-Hasler, S., Suter, M., \u0026amp; Kahmen, A. (2021). Timing of drought in the growing season and strong legacy effects determine the annual productivity of temperate grasses in a changing climate. \u003cem\u003eBiogeosciences\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(2), 585\u0026ndash;604. https://doi.org/10.5194/bg-18-585-2021\u003c/li\u003e\n\u003cli\u003eHansen, M. C., Potapov, P. V., Moore, R., Hancher, M., Turubanova, S. A., Tyukavina, A., Thau, D., Stehman, S. V., Goetz, S. J., Loveland, T. R., Kommareddy, A., Egorov, A., Chini, L., Justice, C. O., \u0026amp; Townshend, J. R. G. (2013). High-Resolution Global Maps of 21st-Century Forest Cover Change. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e342\u003c/em\u003e(6160), 850\u0026ndash;853. https://doi.org/10.1126/science.1244693\u003c/li\u003e\n\u003cli\u003eHe, B., Xie, X., \u0026amp; Guo, L. (2023). A Shift From Temperature to Water as the Primary Driver for Interannual Variability of the Tropical Carbon Cycle. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e, \u003cem\u003e50\u003c/em\u003e(15), e2023GL102812. https://doi.org/10.1029/2023GL102812\u003c/li\u003e\n\u003cli\u003eHuete, A., Didan, K., Miura, T., Rodriguez, E. P., Gao, X., \u0026amp; Ferreira, L. G. (2002). Overview of the radiometric and biophysical performance of the MODIS vegetation indices. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e, \u003cem\u003e83\u003c/em\u003e(1\u0026ndash;2), 195\u0026ndash;213. https://doi.org/10.1016/S0034-4257(02)00096-2\u003c/li\u003e\n\u003cli\u003eLi, X., Wigneron, J.-P., Frappart, F., Fan, L., Ciais, P., Fensholt, R., Entekhabi, D., Brandt, M., Konings, A. G., Liu, X., Wang, M., Al-Yaari, A., \u0026amp; Moisy, C. (2021). Global-scale assessment and inter-comparison of recently developed/reprocessed microwave satellite vegetation optical depth products. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e, \u003cem\u003e253\u003c/em\u003e, 112208. https://doi.org/10.1016/j.rse.2020.112208\u003c/li\u003e\n\u003cli\u003eLi, X., Wigneron, J.-P., Frappart, F., Lannoy, G. D., Fan, L., Zhao, T., Gao, L., Tao, S., Ma, H., Peng, Z., Liu, X., Wang, H., Wang, M., Moisy, C., \u0026amp; Ciais, P. (2022). The first global soil moisture and vegetation optical depth product retrieved from fused SMOS and SMAP L-band observations. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e, \u003cem\u003e282\u003c/em\u003e, 113272. https://doi.org/10.1016/j.rse.2022.113272\u003c/li\u003e\n\u003cli\u003eLiang, J., Crowther, T. W., Picard, N., Wiser, S., Zhou, M., Alberti, G., Schulze, E.-D., McGuire, A. D., Bozzato, F., Pretzsch, H., de-Miguel, S., Paquette, A., H\u0026eacute;rault, B., Scherer-Lorenzen, M., Barrett, C. B., Glick, H. B., Hengeveld, G. M., Nabuurs, G.-J., Pfautsch, S., \u0026hellip; Reich, P. B. (2016). Positive biodiversity-productivity relationship predominant in global forests. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e354\u003c/em\u003e(6309), aaf8957. https://doi.org/10.1126/science.aaf8957\u003c/li\u003e\n\u003cli\u003eLiu, J., Bowman, K. W., Schimel, D. S., Parazoo, N. C., Jiang, Z., Lee, M., Bloom, A. A., Wunch, D., Frankenberg, C., Sun, Y., O\u0026rsquo;Dell, C. W., Gurney, K. R., Menemenlis, D., Gierach, M., Crisp, D., \u0026amp; Eldering, A. (2017). Contrasting carbon cycle responses of the tropical continents to the 2015\u0026ndash;2016 El Ni\u0026ntilde;o. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e358\u003c/em\u003e(6360), eaam5690. https://doi.org/10.1126/science.aam5690\u003c/li\u003e\n\u003cli\u003eLiu, L., Chen, X., Ciais, P., Yuan, W., Maignan, F., Wu, J., Piao, S., Wang, Y., Wigneron, J., Fan, L., Gentine, P., Yang, X., Gong, F., Liu, H., Wang, C., Tang, X., Yang, H., Ye, Q., He, B., \u0026hellip; Su, Y. (2022). Tropical tall forests are more sensitive and vulnerable to drought than short forests. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(4), 1583\u0026ndash;1595. https://doi.org/10.1111/gcb.16017\u003c/li\u003e\n\u003cli\u003eLiu, Z., Zhu, J., Xia, J., \u0026amp; Huang, K. (2023). Declining resistance of vegetation productivity to droughts across global biomes. \u003cem\u003eAgricultural and Forest Meteorology\u003c/em\u003e, \u003cem\u003e340\u003c/em\u003e, 109602. https://doi.org/10.1016/j.agrformet.2023.109602\u003c/li\u003e\n\u003cli\u003eLoewenstein, N. J., \u0026amp; Pallardy, S. G. (1998). Drought tolerance, xylem sap abscisic acid and stomatal conductance during soil drying: A comparison of young plants of four temperate deciduous angiosperms. \u003cem\u003eTree Physiology\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e(7), 421\u0026ndash;430. https://doi.org/10.1093/treephys/18.7.421\u003c/li\u003e\n\u003cli\u003eMarengo, J. A., Tomasella, J., Alves, L. M., Soares, W. R., \u0026amp; Rodriguez, D. A. (2011). The drought of 2010 in the context of historical droughts in the Amazon region: DROUGHT AMAZON 2010. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e(12), n/a-n/a. https://doi.org/10.1029/2011GL047436\u003c/li\u003e\n\u003cli\u003eSarris, D., Christodoulakis, D., \u0026amp; K\u0026ouml;rner, C. (2007). Recent decline in precipitation and tree growth in the eastern Mediterranean. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(6), 1187\u0026ndash;1200. https://doi.org/10.1111/j.1365-2486.2007.01348.x\u003c/li\u003e\n\u003cli\u003eSchwalm, C. R., Anderegg, W. R. L., Michalak, A. M., Fisher, J. B., Biondi, F., Koch, G., Litvak, M., Ogle, K., Shaw, J. D., Wolf, A., Huntzinger, D. N., Schaefer, K., Cook, R., Wei, Y., Fang, Y., Hayes, D., Huang, M., Jain, A., \u0026amp; Tian, H. (2017). Global patterns of drought recovery. \u003cem\u003eNature\u003c/em\u003e, \u003cem\u003e548\u003c/em\u003e(7666), 202\u0026ndash;205. https://doi.org/10.1038/nature23021\u003c/li\u003e\n\u003cli\u003eTaeger, S., Zang, C., Liesebach, M., Schneck, V., \u0026amp; Menzel, A. (2013). Impact of climate and drought events on the growth of Scots pine (Pinus sylvestris L.) provenances. \u003cem\u003eForest Ecology and Management\u003c/em\u003e, \u003cem\u003e307\u003c/em\u003e, 30\u0026ndash;42. https://doi.org/10.1016/j.foreco.2013.06.053\u003c/li\u003e\n\u003cli\u003eTian, F., Brandt, M., Liu, Y. Y., Rasmussen, K., \u0026amp; Fensholt, R. (2017). Mapping gains and losses in woody vegetation across global tropical drylands. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e(4), 1748\u0026ndash;1760. https://doi.org/10.1111/gcb.13464\u003c/li\u003e\n\u003cli\u003eVicente-Serrano, S. M., Beguer\u0026iacute;a, S., \u0026amp; L\u0026oacute;pez-Moreno, J. I. (2010). A Multiscalar Drought Index Sensitive to Global Warming: The Standardized Precipitation Evapotranspiration Index. \u003cem\u003eJournal of Climate\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e(7), 1696\u0026ndash;1718. https://doi.org/10.1175/2009JCLI2909.1\u003c/li\u003e\n\u003cli\u003eVicente-Serrano, S. M., Gouveia, C., Camarero, J. J., Beguer\u0026iacute;a, S., Trigo, R., L\u0026oacute;pez-Moreno, J. I., Azor\u0026iacute;n-Molina, C., Pasho, E., Lorenzo-Lacruz, J., Revuelto, J., Mor\u0026aacute;n-Tejeda, E., \u0026amp; Sanchez-Lorenzo, A. (2013). Response of vegetation to drought time-scales across global land biomes. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e110\u003c/em\u003e(1), 52\u0026ndash;57. https://doi.org/10.1073/pnas.1207068110\u003c/li\u003e\n\u003cli\u003eWalther, G.-R., Post, E., Convey, P., Menzel, A., Parmesan, C., Beebee, T. J. C., Fromentin, J.-M., Hoegh-Guldberg, O., \u0026amp; Bairlein, F. (2002). Ecological responses to recent climate change. \u003cem\u003eNature\u003c/em\u003e, \u003cem\u003e416\u003c/em\u003e(6879), 389\u0026ndash;395. https://doi.org/10.1038/416389a\u003c/li\u003e\n\u003cli\u003eWang, H., Wigneron, J.-P., Ciais, P., Yao, Y., Fan, L., Liu, X., Li, X., Green, J. K., Tian, F., Tao, S., Li, W., Frappart, F., Albergel, C., Wang, M., \u0026amp; Li, S. (2023). Seasonal variations in vegetation water content retrieved from microwave remote sensing over Amazon intact forests. \u003cem\u003eRemote Sensing of Environment\u003c/em\u003e, \u003cem\u003e285\u003c/em\u003e, 113409. https://doi.org/10.1016/j.rse.2022.113409\u003c/li\u003e\n\u003cli\u003eWigneron, J.-P., Ciais, P., Li, X., Josep G, C., Tian, F., Wang, H., Bastos, A., Gatica, G., Kashyap, R., Liu, X., Sitch, S., Tao, S., Xiao, X., Carlo Espinoza, J., Frappart, F., Li, W., Qin, Y., Truchis De, T., \u0026amp; Fensholt, R. (2024). Global carbon balance of the forest: Satellite-based L-VOD results over the last decade and perspectives. \u003cem\u003eFrontiers in Remote Sensing\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e. https://doi.org/10.3389/frsen.2024.1338618\u003c/li\u003e\n\u003cli\u003eWigneron, J.-P., Fan, L., Ciais, P., Bastos, A., Brandt, M., Chave, J., Saatchi, S., Baccini, A., \u0026amp; Fensholt, R. (2020). Tropical forests did not recover from the strong 2015\u0026ndash;2016 El Ni\u0026ntilde;o event. \u003cem\u003eScience Advances\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(6), eaay4603. https://doi.org/10.1126/sciadv.aay4603\u003c/li\u003e\n\u003cli\u003eYao, Y., Liu, Y., Zhou, S., Song, J., \u0026amp; Fu, B. (2023). Soil moisture determines the recovery time of ecosystems from drought. \u003cem\u003eGlobal Change Biology\u003c/em\u003e, \u003cem\u003e29\u003c/em\u003e(13), 3562\u0026ndash;3574. https://doi.org/10.1111/gcb.16620\u003c/li\u003e\n\u003cli\u003eYuan, W., Zheng, Y., Piao, S., Ciais, P., Lombardozzi, D., Wang, Y., Ryu, Y., Chen, G., Dong, W., Hu, Z., Jain, A. K., Jiang, C., Kato, E., Li, S., Lienert, S., Liu, S., Nabel, J. E. M. S., Qin, Z., Quine, T., \u0026hellip; Yang, S. (2019). Increased atmospheric vapor pressure deficit reduces global vegetation growth. \u003cem\u003eScience Advances\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(8), eaax1396. https://doi.org/10.1126/sciadv.aax1396\u003c/li\u003e\n\u003cli\u003eZhang, Z., Babst, F., Bellassen, V., Frank, D., Launois, T., Tan, K., Ciais, P., \u0026amp; Poulter, B. (2018). Converging Climate Sensitivities of European Forests Between Observed Radial Tree Growth and Vegetation Models. \u003cem\u003eEcosystems\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(3), 410\u0026ndash;425. https://doi.org/10.1007/s10021-017-0157-5\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Drought, Vegetation recovery, Tropical forest, Vegetation optical depth","lastPublishedDoi":"10.21203/rs.3.rs-4464016/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4464016/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eTropical rainforests are crucial for Earth's health, but climate change is making severe droughts more frequent. The 2015\u0026ndash;2016 El Ni\u0026ntilde;o-induced drought caused significant biomass loss, yet the recovery duration of different vegetation components (woody parts, upper canopies, and leaves) remains unknown. This study employed satellite remote sensing data of L-band Vegetation Optical Depth (L-VOD), X-band VOD (X-VOD), and Enhanced Vegetation Index (EVI) from 2010 to 2022, characterized by having different sensitivities to the different vegetation components, to examine the recovery of these components in the tropical evergreen broadleaf forest (EBF) regions during the 2015\u0026ndash;2016 El Ni\u0026ntilde;o-induced drought. Results showed that the woody component had the slowest recovery, particularly in Africa, which took longer to return to pre-drought conditions than South America. Key factors influencing recovery included drought severity, moisture-related climatic conditions (i.e., VPD, precipitation, and soil moisture), and seasonal variations. Moreover, the woody component of the EBF in South America showed less impact from drought, benefitted from more favorable moisture-related climatic conditions (e.g., more precipitation and lower VPD), and experienced higher seasonal variation in monthly temperature and precipitation, resulting in a faster recovery than that observed in Africa.\u003c/p\u003e","manuscriptTitle":"Woody component of tropical rainforest recovers slower from drought than the upper canopy layer and leaves","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-11 09:28:18","doi":"10.21203/rs.3.rs-4464016/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"communications-earth-and-environment","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"commsenv","sideBox":"Learn more about [Communications Earth and Environment](https://www.nature.com/commsenv/)","snPcode":"","submissionUrl":"","title":"Communications Earth \u0026 Environment","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Communications Series","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"022bb24a-7394-4cc5-b964-4ba3e2cbad18","owner":[],"postedDate":"June 11th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-11-21T08:08:49+00:00","versionOfRecord":{"articleIdentity":"rs-4464016","link":"https://doi.org/10.1038/s43247-024-01892-9","journal":{"identity":"communications-earth-and-environment","isVorOnly":false,"title":"Communications Earth \u0026 Environment"},"publishedOn":"2024-11-20 05:00:00","publishedOnDateReadable":"November 20th, 2024"},"versionCreatedAt":"2024-06-11 09:28:18","video":"","vorDoi":"10.1038/s43247-024-01892-9","vorDoiUrl":"https://doi.org/10.1038/s43247-024-01892-9","workflowStages":[]},"version":"v1","identity":"rs-4464016","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4464016","identity":"rs-4464016","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}