A 25-year study assessing the subtropical karst terrain of the Wuling Mountains in Guizhou Province, China, reveals that vegetation net primary productivity increased alongside significant ecological restoration efforts, with climate factors and anthropogenic undertakings exerting nearly equal influence on overall productivity variations. Researchers evaluated yearly satellite observations and interpolated meteorological records between 2000 and 2024 to trace how vegetation carbon capture responded across fragile carbonate rock formations compared to neighboring non-karst zones. Net primary productivity, which denotes the quantity of organic carbon captured by plants through photosynthesis after subtracting their own autotrophic respiration, showed an overall upward trend across 71.46 percent of the study area. When analyzing the drivers of these productivity fluctuations, human activities accounted for an average relative contribution of 51.32 percent, while climatic elements contributed 48.68 percent, demonstrating a balanced dynamic in driving vegetation growth and recovery over the quarter-century record.
Karst ecosystems represent some of the most environmentally sensitive terrains globally due to their slow rates of soil development, weak capacities for water and nutrient retention, and rapid surface-to-groundwater infiltration. In the Wuling Mountains study domain, which spans 31,400 square kilometers and constitutes 18.30 percent of Guizhou Province, karst formations account for more than 75 percent of the landscape. The hydrogeological architecture of carbonate bedrock causes rapid percolation of rainwater into subterranean fissure systems, which decouples vegetation water access from annual rainfall totals and renders carbon sequestration uniquely vulnerable. Non-karst regions within the study area possess thicker, more continuous soil profiles averaging between 60 and 100 centimeters, whereas karst zones feature shallow soil profiles typically ranging from 5 to 50 centimeters. This geological distinction constrains root moisture availability, resulting in plant available water capacity in karst settings that is 20 to 30 percent lower than in adjacent non-karst lands.
To capture regional carbon dynamics at high spatial fidelity, the investigation obtained annual net primary productivity data from NASA’s gap-filled MOD17A3HGF product at 500-meter resolution, which estimates plant production based on a light use efficiency framework integrating remote sensing and meteorological variables. Topographic variables were established using a 250-meter digital elevation model from the Chinese Academy of Sciences, while annual land cover categories were derived from Landsat-based China Land Cover Dataset imagery at 30-meter resolution. Meteorological records encompassing annual precipitation, mean temperature, and total solar radiation were compiled from 49 weather stations across Guizhou Province. The station data were spatially interpolated to a matching 250-meter grid using the cubic thin plate spline technique, which accounts for elevation fluctuations across complex mountainous topography. Leave-one-out cross-validation confirmed high interpolation accuracy, yielding coefficient of determination values of 0.98 for temperature, 0.97 for precipitation, and 0.94 for solar radiation.
Long-term trajectory changes in vegetation productivity were determined using Theil-Sen median trend analysis combined with the non-parametric Mann-Kendall test at the 95 percent confidence level. To evaluate climatic relationships, second-order partial correlation analyses were computed pixel by pixel between productivity and each individual climate factor while holding the other two constant. The researchers isolated anthropogenic impacts from climate fluctuations using residual trend analysis, known as RESTREND. This analytical method determines the linear regression slope between productivity and meteorological parameters to estimate climate-driven productivity trends, subtracting this modeled value from the total observed productivity slope. The remaining unexplained residual trend is ascribed to human activities, enabling the separation of vegetative recovery and deterioration into six distinct scenarios depending on the mathematical sign of observed productivity, climatic influence, and anthropogenic contribution.
Across the entire 2000 to 2024 monitoring timeline, annual average productivity reached a multi-year mean of 740.55 grams of carbon per square meter per year for the entire study area, exhibiting a cumulative increase of 3.32 percent. The multi-year mean productivity within karst areas stood at 742.43 grams of carbon per square meter per year, slightly higher than the 736.24 grams recorded in non-karst sectors. However, karst lands exhibited greater structural divergence: they accounted for a higher proportion of severely low productivity zones below 300 grams of carbon per square meter per year, which made up 2.17 percent of karst lands compared to only 0.95 percent in non-karst terrain. A notable multi-year productivity collapse occurred between 2010 and 2012, triggered by an intense freezing disaster in early 2010. During the winter of 2010, average temperatures plummeted to 2.09 degrees Celsius—2.70 degrees lower than the multi-year winter mean of 4.79 degrees—causing productivity in 2010 to drop 7.43 percent below the multi-year baseline before undergoing a three-year recovery that culminated in a peak in 2013.
A quantitative comparison across four specific environmental parameters—soil depth, plant available water capacity, fractional vegetation cover representing ecosystem resilience, and the Karst Remote Sensing Ecological Index reflecting restoration efficacy—demonstrated that karst habitats consistently registered lower average values than non-karst habitats. Soil depth was 18.51 percent lower on average in karst regions, while plant available water capacity was 13.26 percent lower, with more than 81 percent of karst land classified into very low or low water retention grades. Fractional vegetation cover and composite ecological quality indices were 3.36 percent and 8.69 percent lower in karst regions, respectively. Furthermore, productivity exhibited a unimodal curve across gradients of soil thickness and ecological condition, peaking at moderate grades. In very low condition classes, karst plants sustained higher baseline productivity than non-karst plants due to deep root penetration into bedrock fractures and adaptation to thin soils, but intermediate limitations restricted karst vegetation productivity more severely.
Evaluating vegetation types demonstrated that forests sustained the highest mean annual carbon capture at 774.14 grams of carbon per square meter per year, accompanied by the fastest growth rate of 4.23 percent. Shrubs and cropland followed with multi-year averages of 745.94 and 733.47 grams of carbon per square meter per year, expanding at annual rates of 3.43 percent and 3.50 percent, respectively. In contrast, grasslands averaged 667.92 grams of carbon per square meter per year and experienced persistent decline after 2013, with grassland surface area contracting by 75.53 percent between 2000 and 2024. In terms of spatial trends, 55.06 percent of karst terrain exhibited significant productivity gains, outpacing the 48.96 percent observed in non-karst lands. However, non-karst regions maintained greater overall environmental stability, showing an unchanged trend across 26.34 percent of its extent and a productivity degradation rate of only 6.96 percent, compared to 10.28 percent degradation observed in the more fragile karst zone.
Partial correlation analysis demonstrated that temperature served as the predominant climatic factor regulating vegetation carbon dynamics across the region. Among the unstable areas showing notable productivity trends, 66.19 percent exhibited a statistically significant positive correlation with temperature, reflecting how warming in subtropical mountain environments lengthens the vegetative growing season and accelerates enzymatic photosynthesis. Conversely, precipitation correlations were non-significant across 86.57 percent of the unstable landscape, with an average correlation coefficient of negative 0.20, confirming that annual rainfall volume does not govern plant-accessible moisture due to rapid infiltration into subterranean karst aquifers. Solar radiation displayed weak positive associations that were statistically non-significant across 74.26 percent of the territory, a consequence of persistent cloud cover, frequent fog, and collinearity with temperature and moisture regimes characteristic of Guizhou’s montane climate.
In the scenario analysis of RESTREND outputs, synergistic recovery involving both human activities and favorable climate conditions emerged as the dominant dynamic, covering 37.65 percent of the landscape across all vegetation categories. Human intervention played an essential constructive role in greening, supported by state initiatives including the Grain for Green Program, the Natural Forest Protection Project, and the Comprehensive Control Project of Karst Rocky Desertification. Over the 25-year study timeframe, regional forest cover grew from 58.76 percent to 65.59 percent while agricultural acreage receded from 37.19 percent to 32.21 percent. In the strictly conserved Fanjing Mountain National Nature Reserve, primeval forest protection maintained high, stable productivity with negligible degradation. Across areas where vegetation restored successfully, climate factors accounted for 50.05 percent and human endeavors contributed 49.95 percent of the recovery rate.
Conversely, human activities acted as the primary driver of ecological deterioration in degrading patches, accounting for an average contribution rate of 69.31 percent—more than double the 30.69 percent contribution attributed to climate change. Areas experiencing human-driven degradation were characterized by fragmented urban expansion, where impervious built-up land expanded from 0.15 percent in 2000 to 0.75 percent in 2024, directly destroying vegetative cover. Intensive agriculture on steep hillsides and overgrazing on delicate pastures drove degradation across 19.37 percent of grasslands and 5.77 percent of croplands under the human-degradation scenario. The authors noted that while 250-meter gridded datasets markedly advance spatial accuracy over prior kilometer-scale studies, spectral confusion remains along microhabitat rock-vegetation boundaries, and linear residual modeling cannot isolate specific socio-economic drivers such as investments or population shifts without subsequent high-resolution trajectory modeling.
Subject of Research: Vegetation net primary productivity responses to climate change and human activities in karst terrain
Article Title: Quantifying the contribution of climate change and human activities to NPP in a typical karst area
Article References: Zhang, S., Xiang, Z., & Zeng, K. (2026). Quantifying the contribution of climate change and human activities to NPP in a typical karst area. Environmental and Sustainability Indicators, 32, Article 101603. https://doi.org/10.1016/j.indic.2026.101603
Image Credits: AI Generated
DOI: 10.1016/j.indic.2026.101603
Keywords: Karst Ecology, Net Primary Productivity, Climate Change, Vegetation Dynamics, Ecological Restoration, RESTREND Analysis, Guizhou Province, Carbon Sequestration, Quantifying, contribution, climate, change
Cite Scienmag News
Violet Maxwell. (November 8, 2026). Study Quantifies Climate and Human Impacts on Karst Vegetation Productivity. Scienmag. https://scienmag.com/study-quantifies-climate-and-human-impacts-on-karst-vegetation-productivity/
Violet Maxwell. "Study Quantifies Climate and Human Impacts on Karst Vegetation Productivity." Scienmag, 8 November 2026, https://scienmag.com/study-quantifies-climate-and-human-impacts-on-karst-vegetation-productivity/. Accessed 20 November 2026.
Violet Maxwell. "Study Quantifies Climate and Human Impacts on Karst Vegetation Productivity." Scienmag. November 8, 2026. https://scienmag.com/study-quantifies-climate-and-human-impacts-on-karst-vegetation-productivity/








