Biocrust-forming cyanobacteria inoculation to restore degraded soils from dryland ecosystems

Abstract

In drylands, the largest earth biome, the coupled action of climate change and rising human pressure are causing accelerated land degradation (Lal, 2015). This negatively affects biodiversity, soil fertility, water availability, and local population wellbeing, being one of the major environmental issues of the 21st century (UNCCD, 2019). For all these reasons, United Nations, in the Agenda 2030 has proposed a specific goal to stop and reverse ongoing land degradation by the ecological rehabilitation of already degraded drylands. However, most attempts to restore drylands ecosystems by using traditional strategies focused on plant cover establishment fails due to the water scarcity, low fertility and high vulnerability to erosion that characterize these areas (Reynolds et al., 2007). Thus, it is necessary to investigate complementary restoration strategies adapted to strong abiotic stress that can contribute to the recovery of these ecosystems in a cost-effective manner. Due to the harsh abiotic conditions imposed in drylands, plants are often restricted to the most favorable position within the landscape. In open and less favourable inter-plant spaces, other lifeforms with lower edaphic and moisture requirements appear, such as poikilohydric communities of cyanobacteria, algae, fungi, bryophytes and lichens, living in close association with soils particles, and known as biological soil crusts or biocrusts (Belnap et al., 2016). By covering the soil surface, biocrusts form an almost continuous live-skin that intercede in numerous key ecosystems processes (Castillo-Monroy and Maestre, 2011; Maestre et al., 2016), positive affecting soil stability and fertility (Mazor et al., 1996), regulating water balance and reducing water and wind erosion (Belnap et al., 2007; Cantón et al., 2014). However, they are very sensitive to physical disturbance (e.g., vehicular traffic or grazing) and climate change (Ferrenberg et al., 2015), which in turn results in a reduction of their ability to provide key ecosystems services (Weber et al., 2016). In addition, once the activity that causes the disturbance ceases their natural recovery, when possible, tends to be very slow (Weber et al., 2016). Within the last decades, several innovative techniques are being developed to assist in the recovery of disturbed biocrust or to induce the formation of new ones, thereby reinstating the lost ecosystem services. From the different techniques already developed, soil inoculation with biocrust-forming cyanobacteria propagules is among the most attractive for several reasons. For example, theses photoautotrophic bacteria are among the first colonizers of soils (Büdel et al., 2016), enhancing soil fertility (Mazor et al., 1996), soil stability (Chamizo et al., 2018) and improving water retention (Colica et al., 2014) due to their capability to fix CO2, and N2. Moreover, cyanobacteria poses several behavioural and physiological adaptations to cope with harsh abiotic conditions, such as dormancy during desiccation or the releasement of sunscreen pigments to protect from UV (Garcia-Pichel and Castenholz, 1991; Rajeev et al., 2013). In addition, they can be isolated from small amount of natural biocrusts, and cultured ex-situ in nutrient media to produce large quantities of inoculum. All these features convert cyanobacteria in a good candidate to be employed for the ecological rehabilitation of degraded zones in drylands. Previous studies under laboratory conditions proved the feasibility of using cyanobacterial inoculation for promoting artificial biocrust formation and improve soil properties (Acea et al., 2003; Malam-Issa et al., 2007; Chamizo et al., 2018). However, this technology, is far for being applied widespread, as the majority of the literature focused on the use of a single species, Microcoleus vaginatus (Hu et al., 2002; Wang et al., 2009; Xie et al., 2007). Cyanobacteria can be very different regarding feasibility to be isolated and cultured, biocrust formation capability, tolerance to environmental stress and effect on soil quality (Rossi et al., 2017). Therefore, a preliminary laboratory screening of the cyanobacterial growth and its effect on soil properties is an essential step to choose among potential inoculants. For example, water scarcity and soil physicochemical properties constitute two of the main factors affecting cyanobacterial survival and colonization under field conditions (Bu et al., 2014; Fernandes et al., 2018), but indoor studies addressing the sensitivity of different species to these factors are scarce. In this sense, the use of native species should be a preferable option because they can be already preadapted to local conditions (Giraldo-Silva et al., 2019a). Also, we avoid substantial modification of indigenous community. For that reason, this technology would benefit from the search and exploitation of new and native desert species with broader ecological aptitudes and functions for resisting field desiccation and colonize soils with differing physicochemical features. Successful large-scale field application of cyanobacteria to combat soil degradation has been only conducted on previously stabilized sand dunes in China (Chen et al., 2006; Wang et al., 2009). In contrast, several soil inoculations with cyanobacteria were attempted at different deserts in western USA, showing poor results (Kubeckova et al., 2003; Faist et al., 2020). This demonstrated that cyanobacteria application in the field is still a challenge because of the high UV radiation and the prolonged droughts that characterize the arid environments and the negative effect of soil and wind erosion, reducing the odds of survival. Thus, different strategies have been proposed to enhance cyanobacterial survival and establishment under field conditions. One possibility consists on improving the capacity of cyanobacteria to face abiotic stress in the field by gradually increasing UV radiation and reducing water availability during culturing (Giraldo-Silva et al., 2019b). The first outdoor tests using that strategy showed positive fitness results in 13 of the 20 cyanobacterial strains tested (Giraldo-Silva et al., 2019b), however the viability of the preconditioned inoculum have not been tested under field conditions yet. Other possibility that has shown promising results for the rehabilitation of mixed biocrusts communities in the field is to implement habitat amelioration procedures (e.g., jute mesh to promote shade; Bowker et al., 2020), but its application to facilitate cyanobacterial biocrust establishment after the inoculation has yet to be tackled. In addition, once biomass is harvested it should be applied onto the target area in the short-term, otherwise it can get contaminated or be affected by pathogens. This fact difficult the restoration actions based on cyanobacterial inoculation, thereby different groups are currently working on the development of new methodologies to enhance inoculum storage, transferral and applicability to the target areas. One of these alternatives is incorporating cyanobacteria cultures into extruded pellets. Buttars et al. (1998) showed that M. vaginatus survived pelletization and successfully escaped from crushed alginate pellets. In contrast, other attempts with the same species using starch as binder agent showed high mortality (Howard and Warren, 1998). Thus, it is necessary to test its application with other species and pellet compounds. Finally, once the cyanobacteria inoculation has been carried out, the monitoring of the induced biocrust is crucial in order to assess the restoration success and to identify potential factors that constrains inoculum viability. Generally, this is done by measuring induced biocrust coverage or chlorophyll a content. However, laboratory procedures to determine chlorophyll a content are costly and time consuming, and the disturbance of the inoculated crust during samples collection is unavoidable. Thus, the development of an indirect and non-destructive methodology for chlorophyll a quantification in biocrusts is of high interest. In this sense, soil surface reflectance shows high potential to be used for chlorophyll a estimation in biocrust as it is already done for plants (Haboudane et al., 2002), however this methodology has not been previously tested and refined for natural or artificial biocrust communities so far. Under this framework, the main objective of this thesis is to test the potential of native biocrustforming cyanobacteria inoculation to induce the formation of a new biocrust able to enhance soil physicochemical properties in drylands’ degraded soils. To achieve it, firstly, a microcosm experiment was set to test the potential of three native nitrogen-fixing cyanobacteria strains, Nostoc commune, and the non-previously tested Scytonema hyalinum and Tolypothrix distorta, each strain alone and in a consortium (equal mix), as inoculants to restore soil functions. Cyanobacterial inoculation was conducted over three soils with different degrees of soil development from semiarid ecosystems of the province of Almería (SE Spain). Our results revealed that native cyanobacteria inoculation can bring to the formation of an artificial biocrust that enhance soil properties related to soil fertility in a short period. From the strains tested, N. commune and the Consortium showed the best results (CHAPTER I). This experiment was repeated simulating hydration regimes that corresponded to wet and dry hydrological years in the origin areas, in order to identify the cyanobacterial inoculant that best performed under water stress conditions. Surprisingly, similar biocrust development and improvement in soil edaphic conditions were observed under both hydration regimes for all inoculation treatments, suggesting that water availability might not be as important for cyanobacterial biocrust formation as previously reported. During this experiment, the performance of the well-known desiccation-tolerant Nostoc commune was remarkable, showing a greater capacity to growth under water-restrictive scenarios than all other species and being a good candidate for restoring arid degraded areas (CHAPTER II). Afterwards, the viability of the strains was evaluated under field conditions by inoculating a consortium of them on soils from our three study areas. To achieve it, previously, th culturing of each species was carried out in photobiorreactors of 100L using media made with fertilizers, reducing the overall cost of the biomass production (Roncero-Ramos et al., 2019). The direct field application of this indigenous cyanobacterial consortium (N. commune, S. hyalinum and T. distorta) did not significantly facilitated the formation of a new biocrust, as similar values of chlorophyll a, chlorophyll spectral absorption and albedo were found in inoculated and control plots 2 years after inoculation, although soil organic carbon content increased. Thus, a second experiment was carried out in the field to test the effect of inoculum preconditioning and the use of habitat amelioration techniques on the inoculation success. Moreover, previously, the cultures were progressively pre-conditioned to decreasing water supplies before inoculation. However, 6 months after inoculating the hardened inoculum, the similar results were shown by non-conditioned and conditioned plots. Afterwards, the cover of the plots with a plastic fiber grid and a recycled vegetal fiber mesh was evaluated. The combination of nonconditioned cyanobacteria covered with the vegetal mesh resulted in higher colonization, chlorophyll a content, deeper chlorophyll a spectral absorption peaks and lower albedo than uncovered plots 6 months after inoculating (CHAPTER III). In parallel, with the aim of developing a proficient technology to enhance inoculum storage and applicability, we evaluated the survival and establishment of cyanobacteria encapsulation in pellets. Thus, two representative N-fixing genera (Nostoc and Scytonema), a non-heterocystous filamentous genus (Leptolyngbya), and a consortium (equal mix) of all strains were encapsulated in pellets composed of sand and bentonite and incubated on soils from three degraded sites in Australia. This pelletization tests showed that pellets can dissolve completely and spread out in all treatments tested. From the different species, Scytonema and the Consortium showed the best results, with higher biomass than Nostoc and Leptolyngbia at the end of the incubation period. Also, the storage of the pellets for 30 days produced a reduction in chlorophyll a content in all treatments, although at least approximately 50% remained (CHAPTER IV). Finally, the potential of soil surface reflectance measurements for non-destructive monitoring of induced biocrusts was studied. To do this I explored the relationship between different spectral traits of a wide range of biocrusts samples and their chlorophyll a content. This spectral analysis revealed that spectral transformation such as continuum removal and first derivate of reflectance, as well as, normalised band ratios and standard hyperspectral and broad band indices can be used for general indirect chlorophyll quantification in biocrusts. However, such approaches need to be adapted to each specific biocrust type. Interestingly, we found that the need for a specific calibration for each crust type can be sorted out by the combination of spectral measurements with non-linear random forest models (CHAPTER V). In summary, the findings of this thesis provide valuable insights to improve soil rehabilitation actions based on the application of this biotechnology on drylands’ soils. First, our laboratory experiments demonstrate the feasibility of using native cyanobacterial consortium to promote the formation and development of new biocrust that improve key properties of different degraded soils from three Mediterranean ecosystems. Moreover, their application over soils with differing physicochemical properties, including fine-textured soils and mine tailings substrates, provides new advances for developing a proficient technology. Also, we demonstrate that N. commune can survive and colonize the soil with very low water availability, showing a good potential to be used as inoculant in water limited ecosystems. However, the attempt to reinstate soils functions under natural field conditions by means of direct soil inoculation with the indigenous consortium showed poor outcomes. The poor results obtained was due to the early detachment of the inoculum from the surface after desiccation, thereby the propagules were probably washed away due to the overland flow and wind. The use of habitat amelioration procedures that reduce abiotic stress and soil instability significantly improved inoculum survival and establishment. However, although promising at local scale, more studies are necessary for the development of habitat amelioration techniques that can be applied at larger scales. On the other hand, our results showed that some cyanobacteria inoculants, specially Scytonema sp. and the Consortium, can be successfully incorporated in extruded pellets composed of commercial bentonite powder and sand. Nevertheless, the pellet storage for 30 days significantly reduced chlorophyll a content in all the inoculation treatments, though at least the 50% of total cyanobacterial biomass survived. Thus, this technology, although promising, need to be further revised and refined in future studies. Finally, this is the first study where biocrust spectral traits are successfully employed for non-destructive estimation of biocrust biomass. This information can be incorporated at a reasonable cost into monitoring programs for the evaluation of biocrusts rehabilitation projects.