Abstract
Net Primary Productivity (NPP) is a key indicator of terrestrial ecosystem functioning and a major regulator of the global carbon cycle. Yet, its interannual fluctuations and seasonal drivers remain unclear. Using Theil-Sen trend analysis, Mann-Kendall test, R/S analysis, and the Geographically and Temporally Weighted Regression (GTWR) model, this study examined the spatiotemporal dynamics and driving mechanisms of seasonal NPP in the Huaihe River Economic Belt from 2010 to 2021. Results show a general upward trend across all seasons, with the strongest increase in winter. Spatially, NPP decreased from southeast to northwest, with coastal high-value areas contracting seasonally and distinct differences between mountainous and hilly regions. Seasonal patterns revealed clear heterogeneity: spring, summer, and autumn were dominated by stability or improvement with localized degradation, while winter displayed a stable north-south differentiation. Soil moisture emerged as the dominant driver, with multiple factors exerting synergistic effects on seasonal NPP dynamics. This study provides scientific insights to support ecological management and the pursuit of carbon neutrality in the Huaihe River Economic Belt.
Subject terms: Ecology, Environmental sciences
Introduction
Global carbon emissions exacerbate the greenhouse effect and have a profound impact on the global ecosystem1.Vegetation, as a core component of the terrestrial biosphere, serves as a natural link between the atmosphere, hydrosphere, and pedosphere2. Its role in maintaining climate stability and regulating carbon balance has garnered extensive academic attention in recent years, becoming one of the central issues in global change research3,4. Vegetation plays a key role in the carbon cycle, and the change of vegetation carbon sequestration capacity is closely related to the structure and function of ecosystems5,6. Net Primary Productivity (NPP) of vegetation, defined as the net amount of carbon produced by plants through photosynthesis minus that consumed through respiration, is a crucial indicator of the productive capacity of vegetation communities7,8. Changes in NPP not only play a significant role in the global carbon cycle, climate regulation, and carbon balance but also enable estimations of Earth’s carrying capacity and assessments of the sustainable development of terrestrial ecosystems9–11.
With the rapid advancement of remote sensing technology and a deepening understanding of ecosystem process mechanisms, the use of remote sensing data and estimation models to monitor, estimate, and analyze the spatiotemporal dynamics of NPP has become the primary approach in NPP research12–15. Specifically, research focuses on two main areas: one is analyzing the spatiotemporal trends and future projections of regional NPP16–19, and the other investigates how climate change and human activities impact vegetation NPP20–23. Due to data scale limitations, past studies have primarily relied on annual NPP data to analyze spatiotemporal evolution and driving response characteristics in study areas24,25. Less attention has been paid to investigating abrupt changes in NPP and factor responses at finer temporal scales, limiting the understanding of dynamic NPP variations under different seasonal conditions. In terms of driving factor response characteristics, the academic community has employed various methods to qualitatively and quantitatively identify major influencing factors, quantify each factor’s contribution, and explore the interrelationships among factors. Research primarily focuses on the impacts of climate change and human activities on NPP26,27.
In November 2018, the National Development and Reform Commission (NDRC) released the Huaihe River Economic Belt Development Plan, marking the elevation of regional coordinated development in the Huaihe River Economic Belt to a national strategy28. The plan emphasizes that the Huaihe River Economic Belt will adhere to the fundamental principles of resource conservation and environmental protection, with one of its strategic positions being a demonstration zone for watershed ecological civilization construction. By 2035, the goal is to establish a beautiful, livable, vibrant and well-ordered ecological economic belt. Located between the Yangtze River Basin and the Yellow River Basin, the Huaihe River Economic Belt spans the Huang-Huai Plain and connects central and eastern China. It serves as a key ecological security barrier for China’s ecological civilization construction and high-quality economic development, while also being an ecologically sensitive zone with high susceptibility to environmental changes. Positioned within China’s north-south transition zone and influenced by monsoon climates, the region exhibits complex and dynamic ecological conditions. In recent years, rapid urbanization in the region has led to significant changes in land use patterns, resulting in varying degrees of disturbance to vegetation coverage and ecosystem functions. Therefore, studying the spatiotemporal evolution characteristics and driving response mechanisms of vegetation NPP in this region is crucial. Such research not only helps in understanding the impacts of climate change and human activities on the carbon sequestration function of ecosystems but also provides scientific support for regional ecological conservation, land use planning, and environmental management.
This study employed the Google Earth Engine (GEE) platform to obtain seasonal-scale NPP data, capturing seasonal dynamics and spatial variations that are often overlooked at the annual scale. By integrating climate, soil, and topographic factors, the Geographically and Temporally Weighted Regression (GTWR) model was applied to quantify the spatiotemporal non-stationarity of driving mechanisms, highlighting the differential effects of influencing factors on seasonal NPP. The results reveal the spatiotemporal evolution of vegetation productivity and differences in carbon sequestration capacity, provide essential support for identifying ecologically fragile areas and high-value carbon sink zones, and offer scientific guidance for optimizing ecological restoration and land use strategies in the Huaihe River Economic Belt, thereby promoting coordinated ecological protection and economic development as well as contributing to carbon neutrality goals.
Materials and methods
Study area
The Huaihe River Economic Belt (HREB) is located in central-eastern China (31
01’–36
13’N, 112
14’–120
54’E) and encompasses the regions traversed by the main stream and primary tributaries of the Huai River, as well as the Yishu-Sishui River system. This area includes 25 prefecture level cities and 4 counties across five provinces: Jiangsu, Shandong, Henan, Anhui, and Hubei. The HREB lies between the Yangtze and Yellow River basins, bridging the Huang-Huai Plain and connecting central and eastern China, positioned within China’s north-south climate transition zone29. In terms of topographic features, the elevation in this region ranges from -11 to 2137 m, and it can be divided into four typical units: the Tongbai-Dabie Mountains in the southwest (elevation 500–2137 m), the Yimeng Mountains in the northeast (elevation 400-1156 m), the Funiu Mountains in the west (elevation 200–800 m), and the Huang-Huai Plain in the central and eastern parts (elevation < 50 m). Plains with elevations below 500 m account for 98.4% of the region’s total area, mainly distributed in the middle and lower reaches of the Huaihe River and the Yishu-Sishui River Basin. The terrain gently slopes from northwest to southeast, with an average slope of less than 5.4
.
In terms of climate, the region is located within the north–south climate transition zone of China, serving as a climatic ecotone between the northern subtropical and warm temperate zones, and exhibits typical monsoonal characteristics. Specifically, the area experiences cold winters, hot summers, and distinct seasonal variations, with complex and highly variable weather systems. Based on annual temperature and precipitation data from 2010 to 2021, the spatial distribution of mean annual temperature and total annual precipitation was derived. The region’s mean annual temperature is 15.6
C, showing a trend of lower temperatures in the southeast and higher in the northwest. Affected by both the East Asian monsoon circulation and orographic uplift effects, annual precipitation exhibits significant spatial heterogeneity, ranging from 497.6 mm to 1,498.3 mm, and generally decreases from south to north.
The Huaihe River Economic Belt is predominantly characterized by cultivated land as its main land use type. As of 2020, the proportions of various land use types in the region were as follows: cultivated land (66.4%), urban and rural construction land (15.6%), forest land (9.2%), water bodies (5.6%), grassland (3.1%), and unused land (0.1%). Influenced by the dual monsoon climate, the region has a high level of vegetation coverage and rich biodiversity, with favorable natural endowments. The major ecosystem types include cultivated vegetation, evergreen coniferous forests, deciduous broadleaf forests, evergreen-deciduous mixed forests, shrubs, and grasslands. Notably, agricultural ecosystems account for up to 98.5% of the region’s vegetation30. As shown in Figs. 1 and 2a–d.
Fig. 1.
Location of the Huaihe River Economic Belt. The map created based on the standard map service system downloaded with the review number GS (2019) 1822 has no modifications to the base map.
Fig. 2.
Spatial distribution of the following factors in the Huaihe River Economic Belt: (a) mean annual temperature (2010–2021), (b) mean annual precipitation (2010–2021), (c) land use types in 2020, and (d) vegetation types in 2020. The map created based on the standard map service system downloaded with the review number GS (2019) 1822 has no modifications to the base map.
Data sources and preprocessing
NPP data
To more accurately analyze the intra-annual variation of vegetation NPP, this study utilized the Google Earth Engine (GEE) platform and employed the MOD17A3HGF annual mean NPP product (
), the MYD17A2H annual mean GPP product (
), and the MYD17A2H 8-day cumulative GPP data (
)31. Based on the following formula, we synthesized an 8-day cumulative vegetation NPP dataset (
).
 |
1 |
Monthly NPP data were synthesized from 
data, with the average values of March, April, and May representing spring NPP; June, July, and August for summer NPP; September, October, and November for autumn NPP; and January, February, and December for winter NPP. This approach yielded seasonal NPP average data from 2010 to 2021.
Auxiliary data
Temperature (TEMP) and precipitation (PRE) data come from the National Tibetan Plateau Data Center32,33, with a spatial resolution of 0.0083
and monthly temporal resolution, derived from 126 standard meteorological stations in China. Evapotranspiration (ET) data is based on the MOD16A2 product, synthesized from 8-day data into monthly intervals, with a spatial resolution of 500 m. Surface net solar radiation downwards (SSRD) and soil moisture (SM) data are sourced from the ERA5-Land monthly data set provided by the European Centre for Medium-Range Weather Forecasts (ECMWF) from 1950 to present34. This product is a next-generation reanalysis data set based on the land surface component of ERA5 reanalysis data, with a spatial resolution of 0.1
and monthly temporal resolution. This study uses SM data at a depth of 0–7 cm. The DEM data is sourced from the GDEM V3 digital elevation model, provided by the Geospatial Data Cloud site, Computer Network Information Center, Chinese Academy of Sciences, with a spatial resolution of 30 m. The data was processed using ArcGIS 10.2, and elevation (EL) and slope (SLOPE) data were extracted using the HREB vector as the boundary. The time series for all data spans from 2010 to 2021, and to ensure data compatibility in calculations, all datasets were resampled to a spatial resolution of 500 m. The data used in this study are listed in Table 1.
Table 1.
Data type and source.
| Variables | Period | Original spatial resolution | Data source |
|---|---|---|---|
| NPP | 2010–2021 | Monthly / 500 m | https://code.earthengine.google.com/ |
| Temperature | 2010–2021 | Monthly / 0.0083
|
https://data.tpdc.ac.cn/ |
| Precipitation | 2010–2021 | Monthly / 0.0083
|
https://data.tpdc.ac.cn/ |
| Surface net solar radiation downwards | 2010–2021 | Monthly / 0.1
|
https://cds.climate.copernicus.eu/ |
| Evapotranspiration | 2010–2021 | Monthly / 0.1
|
https://cds.climate.copernicus.eu/ |
| Soil moisture | 2010–2021 | Monthly / 0.1
|
https://cds.climate.copernicus.eu/ |
| DEM | 2018 | 30 m | https://www.gscloud.cn/ |
Methods of analysis
Analysis of changing trends
The Theil-Sen median trend analysis method, also known as Sen’s slope estimation, is a robust non-parametric statistical approach for trend calculation based on the median slope of time series data35. This study investigates the spatiotemporal trend of seasonal mean NPP data for the Huaihe River Economic Belt (HREB) from 2010 to 2021. The calculation formula is as follows:
 |
2 |
represents the trend factor; if the value is greater than 0, vegetation NPP shows an increasing trend; otherwise, it shows a decreasing trend. The larger the absolute value of 
, the more evident the trend. 
and 
represent the NPP values of time series points i and j, respectively.
The Mann-Kendall (MK) test36,37 is used to perform a significance test on the spatiotemporal trend of seasonal mean NPP in the HREB from 2010 to 2022, to assess the significance level of NPP changes. The calculation formula is as follows:
 |
3 |
 |
4 |
 |
5 |
The 
is the sign function; when 
, 
, and 
, the function values are 1, 0, and -1, respectively. Var() calculates the variance, Z is the statistic of the standardized test, and n is the total number of time series points. Based on seasonal NPP data from 2010 to 2021 in HREB, the Theil-Sen median trend analysis and the Mann-Kendall abrupt change test were used in this study. Here, a positive 
indicates an increasing trend, while a negative 
indicates a decreasing trend. Values of 
, 
, 
indicate significance levels of 
, 
and 
i.e. 
, 
and 
, corresponding to significant and highly significant changes, respectively. The trend classifications–extremely significant increase (ESI), significant increase (SI), insignificant change (IC), stable (ST), significant decrease (SD), and extremely significant decrease (ESD)—are used to depict the seasonal NPP variation patterns of vegetation.
The R/S analysis method was used to determine the persistence characteristics of vegetation NPP change trends by calculating the Hurst (H) index38, which assesses the presence of long-term trends and periodic changes. The calculation formula is as follows:
Based on the seasonal mean NPP data from 2010 to 2022, the time series of each season’s NPP is denoted as 
(
) for each year. For any positive integer m, the mean sequence of this time series is defined as:
 |
6 |
where 
represents the average NPP over m consecutive years starting from year i.
 |
7 |
 |
8 |
 |
9 |
 |
10 |
 |
11 |
H is the Hurst index, ranging between (0, 1). If H > 0.5, it indicates persistence in the time series, suggesting a long-term correlation between consecutive data points. If H < 0.5, it indicates anti-persistence, with an inverse long-term correlation. H = 0.5 suggests that the time series is random, with no long-term correlation. In this study, the Hurst index (H) was calculated through R/S analysis, where H > 0.5, H < 0.5, and H = 0.5 correspond to persistence (P), anti-persistence (AP), and uncertainty (UN).
To investigate the trends and persistence of NPP in the Huaihe River Economic Belt, NPP trend analysis was combined with the Hurst index to derive coupled information on trend and persistence. Six scenarios were defined for convenience: Persistent extremely significant decrease (P-ESD, 1), Persistent significant decrease (P-SD, 2), Persistent insignificant change (P-IC, 3), Persistent significant increase (P-SI, 4), and Persistent extremely significant increase (P-ESI, 5), representing persistent and meaningful changes. An additional scenario, Uncertain Change Trend (UCT, 0), was included for auxiliary discussion (Table 2). These classifications were based on the Hurst index of seasonal vegetation NPP from 2010 to 2022, characterizing the persistence of NPP change trends.
Table 2.
Category of NPP change trend and persistence.
| Change trend | Hurst | Number | Category |
|---|---|---|---|
| ESI | P | 5 | Persistent extremely significant increase (P-ESI) |
| ESI | UN | 0 | Uncertain extremely significant increase (UN-ESI) |
| ESI | AP | 0 | Anti-persistent extremely significant increase (AP-ESI) |
| SI | P | 4 | Persistent significant increase (P-SI) |
| SI | UN | 0 | Uncertain significant increase (UN-SI) |
| SI | AP | 0 | Anti-persistent significant increase (AP-SI) |
| IC | P | 3 | Persistent insignificant change (P-IC) |
| IC | UN | 0 | Uncertain insignificant change (UN-IC) |
| IC | AP | 0 | Anti-persistent insignificant change (AP-IC) |
| ST | P | 0 | Persistent stable (P-ST) |
| ST | UN | 0 | Uncertain stable (UN-ST) |
| ST | AP | 0 | Anti-persistent stable (AP-ST) |
| SD | P | 2 | Persistent significant decrease (P-SD) |
| SD | UN | 0 | Uncertain significant decrease (UN-SD) |
| SD | AP | 0 | Anti-persistent significant decrease (AP-SD) |
| ESD | P | 1 | Persistent extremely significant decrease (P-ESD) |
| ESD | UN | 0 | Uncertain extremely significant decrease (UN-ESD) |
| ESD | AP | 0 | Anti-persistent extremely significant decrease (AP-ESD) |
Spatiotemporal geographically weighted regression (GTWR)
The spatiotemporal geographically weighted regression (GTWR) model is based on the geographically weighted regression (GWR) model by incorporating a temporal dimension. Compared to the GWR model, GTWR better reveals spatiotemporal nonstationarity in data, capturing both spatial and temporal variation trends and improving the understanding of temporal evolution patterns in data, thus improving the analysis of spatiotemporal data39.
This study uses data from 2010 to 2021 to investigate the relationships between NPP and temperature (TEMP), precipitation (PRE), surface net sola radiation downwards (SSRD), evapotranspiration (ET), soil moisture (SM), elevation (ELV) and slope(Slope). All variables were standardized using the Z-Score method and the regression coefficients were calculated. The formula for the spatiotemporal geographically weighted regression (GTWR) model is as follows:
 |
12 |
 |
13 |
 |
14 |
represents the dependent variable of the i-th sample, 
is the observed value of the k-th independent variable at the i-th sample point, 
are the spatial coordinates of sample point i, 
is the time coordinate. 
denotes the regression coefficient of the k-th independent variable at the i-th sample, and 
represents the spatiotemporal intercept for sample point i. 
denotes distance, and h is the bandwidth, which is determined by the Akaike Information Criterion (AIC) method.
Results
Comparison of seasonal vegetation NPP and the spatial distribution of seasonal NPP
Figure 3a–d illustrates that from 2010 to 2021, the seasonal NPP in the HREB exhibited an overall increasing trend, accompanied by clear seasonal differentiation. NPP values ranged from 0.47 to 101.73 (
), with mean values from highest to lowest as follows: summer (15.86 
), spring (14.11 
), autumn (8.63 
), and winter (2.14 
). The variation rates were 15.18% (spring), 8.47% (summer), 13.80% (autumn), and 126.68% (winter), indicating a significant upward trend during the winter season, which is consistent with the region’s annual vegetation coverage characteristics. From the perspective of the rate of change, between 2013 and 2014, the multi-year change rate in spring exhibited significant fluctuations, increasing from 
% in 2013 to 37.84%, showing a trend of first decreasing and then sharply increasing, with a change rate difference of 51.45%. In comparison to other seasons during this period, although the winter showed large fluctuations, the change rate remained positive, indicating a significant and continuous increase in winter vegetation NPP in 2013-2014. The autumn season, similar to spring, showed a trend of first decreasing and then increasing. Summer exhibited the smallest fluctuation, with a steady upward trend. In the period from 2018 to 2021, the change rates of vegetation NPP in spring, summer, and autumn showed a ”V”-shaped pattern, with the lowest points occurring in 2019. Vegetation NPP in the adjacent years was higher than in 2019. In previous studies, we found that this phenomenon was mainly attributed to a significant decrease in precipitation in 2019 compared to neighboring years, which led to water scarcity negatively affecting vegetation growth. However, the change rate in winter during this period showed a slight positive trend, reflecting that the reduction in water availability may have, to some extent, promoted the increase in winter vegetation NPP in the Huaihe River Economic Belt.
Fig. 3.
Comparison of vegetation NPP variation characteristics in the HREB during spring (a), summer (b), autumn (c), and winter (d) from 2010 to 2021.
The spatial distribution of seasonal NPP averages was divided into seven classes: (0,10], (10,20], (20,40], (40,60], (60,80], (80,100], and >100. Figure 4a–d shows that high-value areas were mainly located in the eastern coastal and southern mountainous regions, with the high-value range in the eastern coastal area gradually shrinking with seasonal changes. The southern mountainous region had the widest distribution of high values in summer, and distinct differences in NPP were observed between the slopes and foothills across seasons. Low-value areas were predominantly distributed in the northern and western plains, as well as in the foothills and plains of the southern region.
Fig. 4.
Spatial distribution of seasonal average vegetation NPP in the HREB from 2010 to 2021: spring (a), summer (b), autumn (c) and winter (d). The map created based on the standard map service system downloaded with the review number GS (2019) 1822 has no modifications to the base map.
Variation trend and persistence of seasonal NPP
Overall, the characteristics of persistent change trends show significant spatial distribution differences across seasons, as Figs. 5a–d and 6 show. The proportion of UCT for each season is as follows: 49.85% (spring), 43.10% (summer), 49.18% (autumn), and 11.90% (winter). The change trend characteristics in winter show distinctly different patterns compared to the other three seasons: (1) The spatial proportion of areas with a statistically significant change trend is relatively large, indicating that these regions exhibit consistent seasonal NPP trends over the study period; (2) The regions exhibiting ST and those showing persistent improvement (ESI, SI) display a clear north-south spatial distribution pattern.
Fig. 5.
Spatial distribution of seasonal NPP change trends and persistence in the HREB from 2010 to 2021: spring (a), summer (b), autumn (c) and winter (d). The map created based on the standard map service system downloaded with the review number GS (2019) 1822 has no modifications to the base map.
Fig. 6.
Proportions of seasonal NPP change trend persistence forms (a) and proportions of determinable change trend persistence forms (b) in the HREB.
In the northern and eastern regions, the characteristics of NPP persistence change trends are predominantly P-SI and P-ESI, accounting for 33.52% and 11.84% of the area, respectively. In the central and southern regions, the NPP persistence change trends are characterized by contiguous areas of P-IC, accounting for 42.61% of the area. In the other three seasons, NPP predominantly exhibits an overall trend of widespread stability and improvement, with small areas of degradation. In Taizhou’s northern region, a trend of persistent degradation is observed during spring, summer, and autumn. In spring, the degraded areas form an east-west belt covering the northern parts of Huainan, Chuzhou, and Taizhou, located between 32-33
N, in regions traversed by the southern tributaries of the Huaihe River.
Driver identification of seasonal NPP
Using the GTWR model, 2580 sampling points were globally selected within an ArcGIS fishnet grid. Parameter estimation was performed for seasonal sample points in 2010, 2014, 2018, and 2021, yielding regression coefficients for each influencing factor at the sample points, including TEMP (X1), PRE (X2), SSRD (X3), ET (X4), SM (X5), ELV (X6), and Slope (X7), and their impact on NPP. The simulation fit derived from the GTWR model showed 
values of 0.87 (spring), 0.81 (summer), 0.81 (autumn), and 0.85 (winter), all exceeding 0.80. These results indicate that the seven selected indicators collectively provide strong explanatory power for seasonal NPP, although model performance and applicability vary slightly among seasons. The high R² values demonstrate the reliability of the model, providing a robust basis for further analysis and interpretation of seasonal NPP dynamics.
The average regression coefficients of influencing factors for each season from 2010 to 2021 across the study area indicate that the impact of each factor on vegetation NPP varies by season, as Table 3 shows. SM (X5) is the most significant factor affecting NPP across all seasons, ranked in descending order as summer, autumn, winter, and spring. The regression coefficient in summer is the highest, reaching 5.9327, suggesting peak vegetation growth during this period. The autumn coefficient is 4.4202, slightly lower than in summer, while the coefficients in winter and spring are smaller but still indicate a positive effect. TEMP (X1) and PRE (X2) exert the strongest positive influence on NPP in autumn, while in spring, they show a weaker negative influence. In summer, SSRD (X3) has the most notable negative impact on NPP, with a regression coefficient of -0.0378. ET (X4) positively influences NPP in all seasons, with the lowest regression coefficient in summer at 0.0717. ELV (X6) and Slope (X7) have relatively minor impacts overall, though seasonal differences are evident. Generally, elevation shows the most pronounced positive impact on NPP in summer, whereas slope has the smallest positive impact on NPP during this season.
Table 3.
Average regression coefficients of influencing factors for seasonal NPP from 2010 to 2021.
| Influencing factor | Spring | Summer | Autumn | Winter | Mean |
|---|---|---|---|---|---|
| TEMP (X1) | -0.0310 | 0.1865 | 0.2553 | 0.1702 | 0.1453 |
| PRE (X2) | -0.0130 | 0.0085 | 0.0250 | -0.0271 | -0.0018 |
| SSRD (X3) | 0.0046 | -0.0378 | 0.0736 | -0.0026 | 0.0095 |
| ET (X4) | 0.2178 | 0.0717 | 0.1736 | 0.1223 | 0.1464 |
| SM (X5) | 0.6697 | 5.9327 | 4.4202 | 1.7399 | 3.1907 |
| ELV (X6) | -0.0051 | 0.0062 | 0.0021 | -0.0022 | 0.0003 |
| Slope (X7) | 0.0242 | 0.0003 | 0.0023 | 0.0053 | 0.0081 |
The spatial distribution of average regression coefficients for NPP influencing factors across each season from 2010 to 2021 was obtained through point interpolation, see Fig. 7. The spatial distribution of average regression coefficients shows that TEMP (X1) primarily has a positive impact on NPP, with the proportion of positive influence increasing across seasons from spring to winter: 
, 
, 
, and 
, respectively. This positive influence expands spatially eastward. In the southern mountainous area of Lu’an, negative influence is observed across all seasons. PRE (X2) shows a greater negative influence in spring (
) and winter (
). In spring, the negative influence is concentrated in Henan, Anhui, and Hubei, with the strongest negative impact at the junction of Bozhou, Fuyang, and Huainan in Anhui. In winter, the negative influence region expands from the Anhui border area and, except for the northern part of Shandong and the eastern coastal area of Jiangsu, nearly covers the entire study area, extending northeast into Zhoukou in Henan. Positive influence is widespread in summer (
) and autumn (
), with the most significant impacts in northern and southwestern mountainous and hilly areas of Shandong. In summer, negative influence is only seen in high-altitude areas in western Pingdingshan, Zhumadian, and most of Hubei; in autumn, it is further limited to northern Shandong, northeastern Henan, northwestern Jiangsu, and northern Anhui. SSRD (X3) shows clear seasonal differences in its impact on NPP, primarily positive in spring and autumn. In spring, positive influence covers 
of the area, located in the eastern, northeastern, and southern regions, with the strongest effect in the Tongbai-Dabie Mountains of Lu’an in the south. In autumn, the positive impact increases to 
, covering all areas except eastern Jiangsu and most pronounced in the southwestern mountainous and hilly regions. In summer, negative influence reaches 
, covering the entire study area except the border between Shandong and Jiangsu, with the strongest negative impact in western Shandong. In winter, 
of the area shows a negative influence, as the cold climate limits photosynthesis and metabolic activity, reducing the positive effect of SSRD. ET (X4) is a positive explanatory factor in all seasons, with regression coefficients ranked from high to low in spring, autumn, winter, and summer. Spatially, in spring, ET shows a decreasing trend from southeast to northwest, with high values in coastal and southern mountainous areas. In summer, high values appear in the northeast and southwest, with lower values in the east and northwest. In autumn, high-value areas extend southward from the region’s center, while low-value areas expand from central-western to northwestern regions. In winter, low-value areas are concentrated on the western boundary, and high values are mainly in central Anhui, particularly in the areas connecting Huaibei, Suzhou, Bengbu, Huainan, and Bozhou. SM (X5) has a notably positive influence on NPP, with the proportion of positively correlated areas exceeding negative correlations across all seasons. In spring, SM’s influence on NPP spatially forms a north-south band, decreasing from the center 116
45’ E meridian toward both sides. The positive and negative influence proportions are 
and 
, respectively. Positive influence areas are located in the region’s center, spanning north to south, with high values at the northern and southern ends, while negative influence areas lie in the western mountains and eastern coast. In autumn, SM’s positive influence is the highest (
), covering all areas except eastern Anhui and southwestern Jiangsu, with high values in southern Anhui and northern Shandong. In winter, the proportion of negative influence increases (
), mainly in the northwestern area. ELV (X6) shows a significant negative influence in spring (
) and winter (
), with similar spatial distribution for positive and negative influence. Positive influence areas are in the southern mountainous and hilly regions, with a slightly larger extent in spring than winter. Positive influence areas dominate in summer and autumn, covering 
in summer, especially along the southeastern coast. Slope (X7) has a predominantly positive influence on NPP across all seasons, particularly in spring and winter, where positive influence areas account for 
and 
, respectively.
Fig. 7.
Spatial distribution of average regression coefficients of NPP influencing factors for each season in the HREB from 2010 to 2021:TEMP (a), PRE (b), SSRD (c), ET (d), SM (e), elevation (f) and slope (g). The map created based on the standard map service system downloaded with the review number GS (2019) 1822 has no modifications to the base map.
Discussion
Spatiotemporal differentiation characteristics of seasonal average NPP
From 2010 to 2021, the seasonal differentiation of NPP in the HREB was pronounced, with high values in summer and low values in winter. NPP showed an increasing trend across all seasons, with a particularly notable increase in winter. Spatially, the eastern coastal areas and southern mountainous regions were the main high-value zones for NPP in all four seasons. In the southeastern coastal area, high-value zones were distributed in a strip along eastern Yancheng in Jiangsu Province, gradually decreasing in proportion across the seasons. The high-value zone in the southern mountains was primarily located along the Dabie Mountain range and its branches, spanning Xinyang in Henan Province and Lu’an in Anhui Province. In this region, NPP had the widest distribution in summer, with distinct differences between slopes and foothills. In spring, low-value NPP areas mainly fell within the 10-40 
range, distributed in the northern mountains and plains of Shandong and the plains north of the southern foothills. In summer, low-value NPP areas ranged between 20-40 
, in patches across the northwestern and central plains. In autumn, low-value NPP areas were within the 10-20 
range, concentrated in Heze in Shandong and the plains of Shangqiu, Zhoukou, Luohe, and Zhumadian in Henan Province. In winter, most areas had NPP values within the 0-10 
range, with 10-20 
zones appearing in the southern parts, forming continuous patches in the southern coastal area of Yancheng in Jiangsu, the border area of Zhumadian in Henan and Fuyang in Anhui, and the Dabie Mountains at the southern end of the region. The remaining high-value areas were scattered around the main tributaries of the Huaihe River.
Analysis of persistent seasonal change trends in NPP
The persistent change trends of NPP show clear seasonal differences, with each season predominantly displaying trends of widespread stability and improvement, accompanied by smaller areas of degradation. Winter stands out for its sustained improvement, which contrasts noticeably with the other seasons. The future change trends in spring, summer, and autumn have substantial areas of uncertainty, accounting for 49.85%, 43.10%, and 49.18% of the total area, respectively. In winter, the proportion of areas with a certain change trend reaches 88.10%, with 45.36% showing sustained improvement the highest among all seasons, primarily in the northern part of the region. This indicates an overall increasing trend in winter NPP, and the persistence analysis suggests that future variations are likely to exhibit patterns consistent with those observed historically. In spring, the proportion of degraded areas is relatively high, forming an east-west band along the tributaries south of the Huaihe River. Notably, northern Taizhou shows a persistent degradation trend across spring, summer, and autumn, warranting close attention.
Analysis of seasonal drivers of NPP
The drivers of vegetation NPP vary across seasons. SM (soil moisture) is the most significant factor across all seasons, consistent with previous findings40. SM and PRE exert a notable positive influence on NPP in summer and autumn, while SSRD and ET suppress NPP increase in summer. In summer, high vegetation cover and optimal water-heat conditions maximize photosynthesis rates, and vegetation water demand peaks41. However, excessive SSRD may lead to rapid water evaporation in plants42, and high temperatures accelerate water loss, resulting in significant positive effects of SM and PRE on NPP. In autumn, as the last growing season for vegetation in the HREB, the reduction of summer heat extends the growing period with suitable temperatures, while ample water and sunlight help sustain photosynthesis and metabolic activities. This promotes the accumulation of dry matter in vegetation, allowing nutrient reserves for the approaching winter. In spring, SM shows distinct areas of positive and negative influence. Negative impacts are more pronounced in the eastern coastal and western mountainous areas. The eastern coast, influenced by the ocean, experiences relatively humid conditions, and excess soil moisture compared to the same latitude in western Jiangsu can lead to water accumulation, resulting in poor root growth due to oxygen deficiency43. This suppresses photosynthesis and limits carbon fixation. In the western mountainous regions, slower spring warming and high vegetation cover may cause soil moisture surplus, preventing vegetation roots from fully utilizing the available water. In winter, the positive influence of TEMP has the highest proportion, while the negative influence proportions of SM, SSRD, and ELV are larger. Excessive precipitation, accompanied by increased soil moisture in low-temperature conditions, can easily cause root freezing. High-altitude areas, with relatively abundant forest resources, tend to have greater cold resistance. In contrast, the plains are mainly covered with grasslands and croplands, where the low temperatures associated with higher elevations more significantly inhibit vegetation growth in these areas44. Areas with greater slope have better drainage conditions, especially in spring and winter, where steeper slopes can effectively prevent excessive soil water accumulation and reduce the risk of root freezing in plants45.
Conclusions
Based on the seasonal average NPP data for the Huaihe River Economic Belt from 2010 to 2021, we applied Theil-Sen Median Trend Analysis and Mann-Kendall Test, along with R/S analysis, to understand the spatiotemporal evolution of seasonal NPP changes. Using the GTWR method, we explored the influence and spatial distribution of various factors on NPP across different seasons. The conclusions are as follows: From 2010 to 2021, the seasonal average NPP in the HREB showed a fluctuating upward trend, indicating regional vegetation recovery. Summer had the highest total NPP, while winter exhibited the fastest NPP growth. There were clear seasonal differences in spatial distribution, with high-value areas located in the eastern coastal and southern mountainous regions, and low-value areas in the northwestern and southern plains. NPP in the eastern coastal region gradually decreased with seasonal changes, and there were notable differences in NPP between the mountainous and foothill areas in the south.
The spatial distribution of seasonal NPP change trends shows significant spatiotemporal differentiation. In spring, summer, and autumn, stable or improving trends dominate, with small areas of degradation, while northern Taizhou in Jiangsu Province exhibits a trend of sustained severe degradation. In winter, 35
N serves as a boundary, with a stable NPP trend across the southern part of the region. In the northern and eastern areas, NPP shows trends of mild and significant improvement, displaying a north-south differentiation in spatial distribution.
Soil moisture is the dominant factor influencing NPP in the HREB, with temperature and evapotranspiration as secondary factors, emphasizing the crucial relationship between NPP and water availability. Climate, soil type, topography, and other factors interact to have a synergistic effect on regional NPP, indicating that its impact is multidimensional rather than driven by a single factor.
Overall, the findings of this study underscore the importance of analyzing the dynamic change mechanisms of NPP in the Huaihe River Economic Belt across multiple temporal scales. This approach aids in guiding future vegetation restoration and conservation efforts, providing empirical insights and strategic guidance to promote sustainable coexistence between humans and the natural environment.
Acknowledgements
We would like to acknowledge the reviewers for their helpful comments on this paper.
Author contributions
J.N. and Z.W. conceived the experiment, Z.W., N.A., H.L., F.X. and L.H conducted the experiment, and Z.W., J.N., N.A. and X.Z. analyzed the results. All authors reviewed the manuscript.
Funding
This research was funded by the Spatial Optimal Allocation Model of Regional Land Use Coupled with the Land System Health (252300421290), the National Natural Science Foundation of China (41771438), Postgraduate Education Reform and Quality Improvement Project of Henan Province (HNYJS2020JD14).
Data availability
Data sets generated during the current study are available from the corresponding author on reasonable request.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Friedlingstein, P. et al. Global carbon budget 2022. Earth Syst. Sci. Data14, 4811–4900 (2022). [Google Scholar]
- 2.Li, W. et al. Regulation of the global carbon and water cycles through vegetation structural and physiological dynamics. Environ. Res. Lett.19, 073008 (2024). [Google Scholar]
- 3.Krinner, G. et al. A dynamic global vegetation model for studies of the coupled atmosphere-biosphere system. Glob. Biogeochem. Cycles19 (2005).
- 4.Perez-Harguindeguy, N. et al. Corrigendum to: New handbook for standardised measurement of plant functional traits worldwide. Aust. J. Bot.64, 715–716 (2016). [Google Scholar]
- 5.An, N. et al. Plant size traits are key contributors in the spatial variation of net primary productivity across terrestrial biomes in China. Sci. Tot. Environ.923, 171412 (2024). [Google Scholar]
- 6.Park, E. et al. Advancement of a diagnostic prediction model for spatiotemporal calibration of earth observation data: A case study on projecting forest net primary production in the mid-latitude region. GIScience Remote Sens.61, 2401247. 10.1080/15481603.2024.2401247 (2024). [Google Scholar]
- 7.Field, C. B., Behrenfeld, M. J., Randerson, J. T. & Falkowski, P. Primary production of the biosphere: Integrating terrestrial and oceanic components. Science281, 237–240.
- 8.Chen, Y., Wang, J., Xiong, N., Sun, L. & Xu, J. Impacts of land use changes on net primary productivity in urban agglomerations under multi-scenarios simulation. Remote Sens.14, 1755 (2022). [Google Scholar]
- 9.Zhao, M. & Running, S. W. Drought-induced reduction in global terrestrial net primary production from 2000 through 2009. Science329, 940–943 (2010).
- 10.Michaletz, S. T., Cheng, D., Kerkhoff, A. J. & Enquist, B. J. Convergence of terrestrial plant production across global climate gradients. Nature512, 39–43 (2014). [DOI] [PubMed] [Google Scholar]
- 11.Collalti, A. et al. The sensitivity of the forest carbon budget shifts across processes along with stand development and climate change. Ecol. Appl.29, e01837 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Park, E. et al. Advancement of a diagnostic prediction model for spatiotemporal calibration of earth observation data: A case study on projecting forest net primary production in the mid-latitude region. GISci. Remote Sens.61, 2401247 (2024). [Google Scholar]
- 13.Fischer, S. M., Wang, X. & Huth, A. Distinguishing mature and immature trees allows estimating forest carbon uptake from stand structure. Biogeosciences21, 3305–3319 (2024). [Google Scholar]
- 14.Ghorbani, A., Dadjou, F., Moameri, M. & Biswas, A. Estimating aboveground net primary production (ANPP) using Landsat 8-based indices: A case study from Hir-Neur rangelands, Iran. Rangeland Ecol. Manag.73, 649–657 (2020). [Google Scholar]
- 15.Xu, M. et al. Estimating estuarine primary production using satellite data and machine learning. Int. J. Appl. Earth Obs. Geoinf.110, 102821 (2022). [Google Scholar]
- 16.Zhao, D., Jia, W. & Liu, J. Dynamic changes and driving mechanisms of net primary production (NPP) in a semi-arid region of China. Sustainability15, 11829 (2023). [Google Scholar]
- 17.Li, M., Li, Q. & Xue, M. Spatio-temporal changes of vegetation net primary productivity and its driving factors on the Tibetan Plateau from 1979 to 2018. Atmosphere15, 579 (2024). [Google Scholar]
- 18.Hu, Q., Zhang, J., Xue, H., Wang, J. & Li, A. Spatiotemporal variations in carbon sources and sinks in national park ecosystem and the impact of tourism. Sustainability16, 7895 (2024). [Google Scholar]
- 19.Xu, Q., Li, J., Zhang, S., Yuan, Q. & Ren, P. Spatio-temporal changes and driving mechanisms of vegetation net primary productivity in Xinjiang, China from 2001 to 2022. Land13, 1305 (2024). [Google Scholar]
- 20.Wang, M., Sun, K., Jia, J., Wu, F. & Gao, Y. Climate change drove the decline in Yangtze Estuary net primary production over the past two decades. Environ. Sci. Technol.58, 19305–19314 (2024). [DOI] [PubMed] [Google Scholar]
- 21.Wu, J. et al. Correlation of climate change and human activities with agricultural drought and its impact on the net primary production of winter wheat. J. Hydrol.620, 129504 (2023). [Google Scholar]
- 22.Xu, M., Guo, B. & Zhang, R. Variations of terrestrial net ecosystem productivity in china driven by climate change and human activity from 2010 to 2020. Forests15, 1484 (2024). [Google Scholar]
- 23.Feng, H. et al. The impact of climate change and human activities to vegetation carbon sequestration variation in Sichuan and Chongqing. Environ. Res.238, 117138 (2023). [DOI] [PubMed] [Google Scholar]
- 24.Wang, Z. et al. Spatiotemporal evolution and attribution analysis of grassland npp in the Yellow River Source Region, China. Ecol. Inform.76, 102135 (2023). [Google Scholar]
- 25.Liu, C. et al. Analysis of net primary productivity variation and quantitative assessment of driving forces—A case study of the Yangtze River Basin. Plants12, 3412 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Xu, Y., Lu, Y.-G., Zou, B., Xu, M. & Feng, Y.-X. Unraveling the enigma of npp variation in Chinese vegetation ecosystems: The interplay of climate change and land use change. Sci. Tot. Environ.912, 169023 (2024). [Google Scholar]
- 27.Liu, S., Liu, J. & Shang, S. Attribution of climate change and human activities to spatiotemporal changes of ecological service value in Yunnan Province of China. Glob. Ecol. Conserv.55, e03221 (2024). [Google Scholar]
- 28.Sun, J. & Shuchang, Y. Approaches to promoting high-quality development of Huaihe River Eco-economic Belt. Finance Trade Res.31, 43–48. 10.19337/j.cnki.34-1093/f.2020.03.004 (2020).
- 29.Zhang, B., Niu, J., Liu, J., Jiang, Y. & Li, J. A study on the areal differentiation and zonality of evergreen deciduous broad-leaved mixed forests in the north-south transitional zone of China. Xinyang Normal Univ. (Nat. Sci. Ed.)36, 345–355 (2023). [Google Scholar]
- 30.Huang, J. & Li, Y. Spatial differences and dynamic evolution of urban ecological resilience in the Huai River Ecological Economic Belt. J. Xinyang Normal Univ. (Nat. Sci. Ed.)37, 1–10 (2024).
- 31.Bajic, D., Adzic, D. & Dekić, R. GIS-based spatial-temporal analysis of the biomass production in the republic of Srpska territory. Researchgate21, 53–69 (2017). [Google Scholar]
- 32.Shouzhang, P. 1-km Monthly Mean Temperature Dataset for China (1901–2023). 10.11888/Meteoro.tpdc.270961 (National Tibetan Plateau Data Center , 2024)
- 33.Shouzhang, P. 1-km monthly precipitation dataset for China (1901–2023). National Tibetan Plateau Data Center10.5281/zenodo.3114194 (2024). [Google Scholar]
- 34.Muñoz-Sabater, J. et al. Era5-land: A state-of-the-art global reanalysis dataset for land applications. Earth Syst. Sci. Data13, 4349–4383 (2021). [Google Scholar]
- 35.Sen, P. K. Estimates of the regression coefficient based on Kendall’s tau. J. Am. Stat. Assoc.63, 1379–1389 (1968). [Google Scholar]
- 36.Mann, H. B. Nonparametric tests against trend. Econ. J. Econ. Soc. 245–259 (1945).
- 37.Forthofer, R. Public Program Analysis: A New Categorical Data Approach (Springer, 2012).
- 38.Wang, C., Wang, L., Zhao, W., Zhang, Y. & Liu, Y. Analysis of spatiotemporal change and driving factors of npp in Qilian Mountains from 2000 to 2020. Rangeland Ecol. Manag. (2024).
- 39.Pang, B. et al. Spatial and temporal divergence and driving mechanisms of carbon sinks in terrestrial ecosystems in the middle reaches of the Yangtze River Urban agglomerations during 2008–2020. Ecol. Indic.165, 112205 (2024). [Google Scholar]
- 40.Han, Y., Lu, H. & Qiao, D. Integrated effects of meteorological factors, edaphic moisture, evapotranspiration, and leaf area index on the net primary productivity of winter wheat- summer maize rotation system. Field Crops Res.302, 109080 (2023). [Google Scholar]
- 41.Adeyeri, O. E., Zhou, W., Ndehedehe, C. E. & Wang, X. Global vegetation, moisture, thermal and climate interactions intensify compound extreme events. Sci. Total Environ.912, 169261 (2024). [DOI] [PubMed] [Google Scholar]
- 42.Crabtree, R. et al. A modeling and spatio-temporal analysis framework for monitoring environmental change using npp as an ecosystem indicator. Remote Sens. Environ.113, 1486–1496 (2009). [Google Scholar]
- 43.Zhang, Q., Xiao, F., Niu, H. & Dong, W. Analysis of vegetation index sensitivity to soil moisture in northern china. Chin. J. Ecol. 715–718 (2005).
- 44.Zhang, X. et al. Phenology of forest vegetation and its response to climate change in the Funiu mountains. Dili Xuebao/Acta Geogr. Sin.73, 41–53. 10.11821/dlxb201801004 (2018).
- 45.Chen, S., Ma, M., Wu, S., Tang, Q. & Wen, Z. Topography intensifies variations in the effect of human activities on forest npp across altitude and slope gradients. Environ. Dev.45, 100826 (2023). [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data sets generated during the current study are available from the corresponding author on reasonable request.