Global trends in vegetation fractional cover: Hotspots for change in bare soil and non-photosynthetic vegetation

Highlights

• Global vegetation fractional cover is an indicator of ecosystem health.

• Bare soil cover fraction is increasing across 11 M km² of the land surface.

• Decline in dry vegetation is associated with increases in green vegetation and bare soil.

• Drought and increased livestock populations are associated with bare soil trends.

Abstract

Vegetation Fractional Cover (VFC) is an important indicator of the condition of the terrestrial surface of the Earth. The dynamics of bare soil (BS), non-photosynthetic vegetation (NPV) and photosynthetic vegetation (PV) fractions reveal patterns of growth, senescence, dormancy and regeneration. The dynamics also reveal trends in land cover and land use management across the globe. Analysis of satellite data has indicated increasing greenness of the land surface however the dynamics of the NPV and BS fractions are equally important indicators of many elements of ecosystem function and sustainability. In this study, we provide a comprehensive assessment of trends in VFC for the global terrestrial land surface based on a trend analysis over the period 2001–2018 using the Global Vegetation Fractional Cover Product. Trends were analysed by Mann-Kendell regressions over 18 years for each month of the year. Monthly maps of significant trend areas were aggregated to provide seasonal assessments. Areas and percentage areas of positive and negative trends were compiled for United Nations Subregions, Countries and Ecoregions. Significant negative trends in BS occurred across 20 M km² of the terrestrial land surface. Although the areas of negative trends in BS were largest in China and India, the top 24 countries by area were widely distributed across the world. However, BS increased significantly across more than 11 M km². Eleven of the top 24 ranked countries for area of increasing BS were in Africa. Substantial areas exhibited negative trends in NPV split between China and India where PV was increasing, and drylands in Asia and Africa where BS was increasing. Analysis of covariate factors at country scale inferred an association between significant increase in BS and long-term drought and increases in livestock populations. The positive trends in BS and negative trends in NPV are indicators of current and future risk for soil erosion, habitat change, and reduction in ecosystem health across heavily populated regions of the developing world.

Introduction

Vegetation Fractional Cover (VFC) is an important indicator of the condition and health of the terrestrial surface of the Earth. Excluding permanent snow and ice cover, the terrestrial surface may be described in simple terms by the fractional cover of green leaves which fix carbon through photosynthesis (Photosynthetic Vegetation – PV), non-green vegetative material that is primarily made up of plant structural materials such as cellulose and lignin (Non-Photosynthetic Vegetation – NPV), and bare including soil and rock (Bare Soil – BS). The dynamics of these fractions is determined by seasonal patterns of precipitation and temperature resulting in varied patterns of growth, senescence, dormancy and regeneration across the globe (Foley et al., 2000). They also reflect trends in land cover and land use management (Song et al., 2018; Erb et al., 2018). Analysis of satellite data has indicated increasing greenness of the land surface (Park et al., 2016, Vickers et al., 2016, Xu et al., 2013, Zhou et al., 2001, Zhu et al., 2016). The positive trends in greening detected in the MODIS Leaf Area Index (LAI) product (Chen et al., 2019), which directly correlate to an increase in PV cover fraction, have been attributed to cropland expansion and intensification, CO₂ fertilization and large-scale tree-planting, especially in India and China (Chen et al., 2019, Zhu et al., 2016, Piao et al., 2015, Donohue et al., 2013, Keenan et al., 2016, Cheng et al., 2017; Fensholt and Proud, 2012). However, the dynamics of the NPV and BS fractions are equally important indicators of many elements of ecosystem function and sustainability such as carbon storage Jacinthe et al., 2002), wind and water erosion risk (Imeson and Prinsen, 2004, Borrelli et al., 2017, Li et al., 2020), water infiltration (Ludwig et al., 2005), habitat condition (McNaughton, 1985; Ludwig et al., 2004; Zhang et al., 2013) and wildfire risk (Renier et al., 2015) especially in highly seasonal savannas and grasslands, and semi-arid and arid environments (Huang et al., 2020).

Increasing human populations and urbanization, and changes in dietary preferences have driven expansion in area and intensity of cropping (Rudel et al., 2009, Zalles et al., 2019) and in livestock numbers (Thornton, 2010, Derner et al., 2017, Ritchie, 2017), especially goats in Asia and Africa (Skapetas and Bampidis, 2016). This has greatly increased the intensity of the human footprint across a wide range of ecosystems (Venter et al., 2016) and increased the ongoing concern about global desertification and land degradation (Reynolds et al., 2007, Safriel, 2007, Le et al., 2016, Huang et al., 2020) and threats to biodiversity (Hu et al., 2020). The assessment by Le et al. (2016) using a long-term trend analysis of GIMMS NDVI (Global Inventory Modeling and Mapping Studies Normalized Difference Vegetation Index) between 1982 and 2006 suggested that land degradation “hotspots” covered 29% of global land surface. The impact of this increased intensity has tended to place more pressure on the vegetation and ecosystem function in warmer and drier environments (Modernel et al., 2016, Tian et al., 2017). For example, recent studies have identified higher intensity of methane emissions (largely attributable to increase in livestock populations) from drylands and non-drylands, and an increased share of global methane emissions from developing regions of Africa, Asia and South America (Dangal et al., 2017). Higher grazing intensity can lead to greater levels of bare soil (Augustine et al., 2012, Modernel et al., 2016). Specific pressure points driven by anthropogenic change and climate change have been identified globally (Safriel, 2007; Huang et al., 2020) and for various regions including East Africa (Olson et al., 2004, Olson et al., 2008, Moore et al., 2012), Central Asia (Li et al., 2020) and South America (Santibáñez and Santibáñez, 2007).

The baseline levels and dynamics of the NPV fraction are an indicator of cropland stubble retention (Daughtry, 2001, Daughtry et al., 2005, Daughtry et al., 2006), dry season feed availability for livestock (Wessman et al., 1997), and fuel loads for wildfires (Lambin et al., 2003, Chuvieco et al., 2003; Guerschman et al., 2009). Various methods for detection of total vegetation and soil fractional cover have been explored (e.g., Okin et al., 2013; Song et al., 2017; Yue et al., 2020) and global bare ground changes have been assessed with Landsat Imagery (Ying et al., 2017). However, explicit mechanistic retrieval of the NPV fraction with remote sensing has been lacking until the development of the MODIS Australian Vegetation Fractional Cover Product (AVFCP); Guerschman et al. (2015); Guerschman and Hill (2018). The AVFCP has been used to develop an online tool for monitoring groundcover as an indicator of wind and water erosion risk for Natural Resource Management regions of Australia (Guerschman et al., 2018) and to examine the dynamics of the PV and NPV cover fractions in relation to antecedent rainfall (Guerschman et al., 2020).

A global version of the product, the MODIS Global Vegetation Fractional Cover Product (GVFCP), was recently described (Hill and Guerschman, 2020). In that study the trends in VFC components were explored across World Grassland Types (WGT; Dixon et al., 2014) among Global Livestock Production Systems (GLPS; Robinson et al., 2011). However, the study was limited in geographical extent, examined the trends for a single overall monthly time series from 2001 to 2018, and reported areas of significant trend at Formation and Division level within the WGT class hierarchy. In this study, we aimed to provide a comprehensive assessment of trends in VFC for the global terrestrial land surface excluding permanent and semi-permanent ice and snow with a focus on trends in the NPV and BS fractions.

Therefore, the goals of this study were:

• To detect and document the areas of the terrestrial land surface exhibiting statistically significant long-term positive and negative trends in the PV, NPV and BS cover fractions including comparison between seasons.

• To explore trends at country level and examine associations with drought, livestock density, cropland expansion and population numbers.

• To explore trends at ecoregion level as indicators of land cover change and potential land degradation with a specific focus on trends in NPV and BS.