Effects of climate change and land use intensification on regional biological soil crust cover and composition in southern Africa

Highlights

• Climate change scenarios predict an increase in aridity in South African deserts.

• Biocrust distribution within South African drylands is governed by aridity and livestock pressure.

• More developed biocrust coverage declined as conditions became warmer and drier and grazing pressure increases.

• This resulted in a decrease in NDVI biocrust values as climate became drier.

• Satellite data along aridity gradients are suited to identify climate change effects.

Abstract

Biological soil crusts (biocrusts) form a regular and relevant feature in drylands, as they stabilize the soil, fix nutrients, and influence water cycling. However, biocrust forming organisms have been shown to be dramatically vulnerable to climate and land use change occurring in these regions.

In this study, we used Normalized Difference Vegetation Index (NDVI) data of biocrust-dominated pixels (NDVI_biocrust) obtained from hyperspectral and LANDSAT-7 data to analyse biocrust development over time and to forecast future NDVI_biocrust development under different climate change and livestock density scenarios in southern Africa. We validated these results by analysing the occurrence and composition of biocrusts along a mesoclimatic gradient within the study region.

Our results show that NDVI_biocrust, which reached maximum values of 0.2 and 0.4 in drier and wetter years, respectively, mainly depended on water availability. A predicted decrease in rainfall events according to all future climate scenarios combined with increased temperatures suggested a pronounced decrease in NDVI_biocrust by the end of the 21st century caused by reduced biocrust coverage. Livestock trampling had similar effects and exacerbated the negative impacts of climate change on biocrust coverage and composition. Data assessed in the field concurred with these results, as reduced biocrust cover and a shift from well-developed to early stages of biocrust development were observed along a gradient of decreasing precipitation and increasing temperatures and livestock density.

Our study demonstrates the suitability of multi-temporal series of historical satellite images combined with high-resolution mapping data and Earth system models to identify climate change patterns and their effects on biocrust and vegetation patterns at regional scales.

Introduction

Earth’s drylands (mainly dominated by sensitive forests, steppes and deserts) occupy nearly half of the Earth́s terrestrial surface, and their coverage is expected to increase by the end of the century (Feng and Fu, 2013). Global climate models predict more frequent long-lasting droughts and increased warming within these ecosystems (Huang et al., 2016). This, as well as land use intensification has the potential to produce abrupt changes in multiple ecosystem attributes, thus affecting ecosystem functions, global Earth system functioning and human livelihood (Maestre et al., 2016, Berdugo et al., 2020).

One of the most representative biotic components of dryland regions around the world are biological soil crusts (abbreviated as biocrusts), which are complex communities of photosynthetic organisms, such as cyanobacteria, algae, lichens, and bryophytes, growing together with heterotrophic bacteria, fungi, and archaea within the uppermost millimeters of the soil (Weber et al., 2016). These diminutive communities are estimated to cover ∼12 % of the terrestrial soils (Rodríguez-Caballero et al., 2018a), influencing energy (Couradeau et al., 2016, Rutherford et al., 2017), water (Chamizo et al., 2016, Eldridge et al., 2020), C and N fluxes between the soil and atmosphere (Elbert et al., 2012, Porada et al., 2013, Porada et al., 2014, Porada et al., 2017, Weber et al., 2015, Lenhart et al., 2015). Moreover, they affect biogeochemical cycling within the soil (Delgado-Baquerizo et al., 2013, Maestre et al., 2013), with important effects on soil fertility (Chamizo et al., 2012a), local hydrology (Chamizo et al., 2016, Eldridge et al., 2020), and soil resistance to erosive forces (Belnap et al., 2014). Thus, they provide stable environments and additional resources for vascular vegetation (Luzuriaga et al., 2012, Rodríguez-Caballero et al., 2018b, Havrilla et al., 2019) and the soil fauna (Bamforth, 2008).

Although biocrusts are adapted to aridity, surviving extreme environmental conditions (Pointing and Belnap, 2012), biocrust forming organisms have been shown to be dramatically vulnerable to subtle changes of climatic parameters (Reed et al., 2012, Maestre et al., 2013, Darrouzet-Nardi et al., 2015). According to global change scenarios, their global coverage is expected to decrease dramatically by the end of this century (Rodríguez-Caballero et al., 2018a). Besides this overall reduction in biocrust coverage, climate manipulation experiments also showed that warming, and changes in precipitation patterns (i.e. an increase in the frequency of small water pulses during warm periods) will cause changes in cyanobacteria composition and a reduction of lichen or bryophyte coverage (Maphangwa et al., 2012, Reed et al., 2012, Escolar et al., 2012, García-Pichel et al., 2013, Maestre et al., 2013, Ladrón de Guevara et al., 2014, Ladrón de Guevara et al., 2018), with important implications for ecosystem functioning (Escolar et al., 2012, Maphangwa et al., 2012, Delgado-Baquerizo et al., 2013, Couradeau et al., 2016, Rutherford et al., 2017). Disturbance derived from land use intensification (e.g., trampling) has also been demonstrated to be a relevant threat for biocrusts. In fact, it has similar negative effects as climate change on their coverage and composition (Ferrenberg et al., 2015), causing an increase of soil erosion (Chamizo et al., 2012a, Chamizo et al., 2012b) and modifications of water (Chamizo et al., 2016) and nutrient (Belnap, 1996) cycling in the soil. Thus, both climate change and disturbance caused by land use intensification have major effects on the composition, survival, and distribution of biocrusts. However, especially field experiments on climate change have been limited to very few sites (e.g., Mojave Desert, Colorado Plateau, Iberian Peninsula), that only represent a small fraction of the whole range of ecosystems and climatic conditions under which biocrusts occur. These experimental studies are time-consuming, expensive, and often involve interfering side effects. Passive open-top chambers not only cause a warming inside the chambers, but also cut down the impact of dew and fog (Maphangwa et al., 2012), and warming lamps only heat the uppermost soil region with a downward gradient (Kimball, 2005). Such negative side-effects of field experiments can be avoided by combining experimental results with direct observations in the field or by applying remote sensing technologies (Palmer et al., 2002).

Space-for-time (SFT) substitutions, where spatial climatic and disturbance gradients are used to investigate the response of biocrusts to anticipated future conditions, form one of these options. This method has been frequently applied to analyse plant succession and to model climate change effects on biodiversity and species richness (Blois et al., 2013). In a recent study, it provided valuable data on the long-term effects of temperature and precipitation on the composition and coverage of biocrust communities (García-Pichel et al., 2013). Another methodology is the use of historical and open access high temporal resolution satellite images combined with coverage and composition data along climatic gradients. Such multi-temporal series of satellite images could provide crucial information on biocrust dynamics and response to environmental factors, such as temperature, precipitation, and disturbance (Karnieli et al., 1996, Karnieli, 2003, Burgheimer et al., 2006, Rodríguez-Caballero et al., 2015, Panigada et al., 2019).

The main objective of this paper is to investigate the long-term response of biocrusts to warming, livestock trampling, and changes in precipitation amount and frequency predicted for the Succulent Karoo. The study region is located within an arid winter rainfall biome of southern Africa (Mucina et al. 2006) that comprises a great diversity and unique flora of succulent plants (Mucina et al., 2006, Schmiedel and Jürgens, 1999). It also harbours a variety of different biocrust types that are considered as a critical component of the ecosystem (Büdel et al., 2009). To achieve the objective, we apply a multidisciplinary approach that combine information of the Normalized Difference Vegetation Index (NDVI) obtained form satellite images with field data and climate predictions.

We hypothesize that i) NDVI will reflect the rapid response of biocrusts to variations in climatic factors, and ii) as a consequence, the NDVI_biocrust will be reduced in the future due to a predicted increase in aridity according to the IPCC predictions for the region (Weber et al., 2018). We further suggest that iii) the reduced NDVI_biocrust will be the result of changes in both biocrust coverage and composition.