1 Introduction

Biological soil crusts (BSCs; Fig. 1a) develop when various combinations of diminutive cyanobacteria, algae, lichenized and non-lichenized fungi and/or bryophytes occupy the surface and upper few millimeters of the soil. Historically, they have been referred to as cryptobiotic, cryptogamic, microbiotic, microfloral, microphytic and organogenic crusts. They can be present in a wide range of ecological, successional, and climatic conditions when and where disturbance and/or aridity have resulted in opportunities for colonization. However, they are most prevalent in arid and semiarid ecosystems where vascular plant cover and diversity are characteristically low, leaving large areas available for colonization by some combination of the organisms mentioned above. The diversity and distribution of BSC components in extreme environments is striking. For example, at least 18 species of cyanobacteria have been documented in the soils of Death Valley National Monument in the Mojave Desert, USA, where surface temperatures can reach 88 °C (Durrell 1962). At the opposite end of the temperature spectrum, BSC communities are common in interior Antarctica, where soil temperatures seldom exceed 0 °C (Green and Broady 2001). BSCs are also present in the hyper-arid Atacama Desert of northern Chile (Patzelt et al. 2014), where average annual precipitation, depending on latitude, elevation, and distance from the Pacific coast, can be less than 1 mm.

Fig. 1
figure 1

Biological soil crust communities (BSC) and potential factors leading to atmospheric mixing and dispersal. a A well-developed BSC on the Colorado Plateau in southeastern Utah, USA; b Dust storm in the southwestern USA created by summer monsoon thunder cells (source: https://commons.wikimedia.org/wiki/File:Dust_storm_clouds_gathering.jpg); c a strong vertical vortices or ‘dust devils’ in western North America (source: (https://www.flickr.com/photos/jurvetson/888064641); and d satellite image of a dust storm extending from Saudi Arabia to Turkey in 2011 (source: https://earthobservatory.nasa.gov/images/52305/dust-storm-in-the-middle-east)

Full size image

The ecological roles of BSCs are many and varied and include the collection, accumulation, and cycling of essential airborne and soil nutrients, redistribution of precipitated water, and soil formation and stabilization (Warren 1995; Belnap and Lange 2001; Chen et al. 2009; Weber et al. 2016b). BSCs, and their ecological functions, can be disturbed by a variety of anthropogenic factors, including, but not limited to, livestock trampling (Warren and Eldridge 2001), off-road vehicular traffic (Wilshire 1983; Webb et al. 1988), military operations (Warren 2014), mining (Spröte et al. 2010), and fire (Johansen 2001). In spite of the overall importance of BSCs and the well-documented effects of disturbance on these communities, restoring degraded habitats has received proportionately less attention. While important perspectives on BSCs rehabilitation have been previously summarized (Belnap and Eldridge 2001; Bowker 2007; Chiquoine et al. 2016), the role of aerobiology and passive restoration has not been adequately considered. Here, we provide an up-to-date perspective incorporating principles of aerobiology and passive dispersal into the BSC restoration paradigm. In addition to directing future research, reflection on the broad scope of BSC restoration can improve our perspective of how to effectively manage important dryland regions.

2 Artificial restoration

It may seem intuitive to attempt to restore BSCs by inoculating disturbed sites with crust organisms, but such applications have been relatively rare. St. Clair et al. (1986) inoculated small plots with a soil slurry made by stripping BSCs from intact areas, mixing them with water, and applying them to sites damaged by wildfire. Belnap (1993) stripped crusts from an intact area and used them as a dry inoculant on small plots where the original crust had been removed. The inoculation of soil in petri dishes with dry and slurried inocula, plus additions of water up to five times per week, and sewage sludge, produced limited soil crust re-establishment (Maestre et al. 2006). Bu et al. (2014) inoculated soil in a greenhouse study with BSCs that had been stripped from intact areas in the field in an attempt to accelerate crust restoration. They found that frequent watering of the crusts in that setting enhanced growth, but field trials were not conducted. In the Mojave Desert, USA, a somewhat similar approach was attempted using crusts composed of cyanobacteria, lichens, and bryophytes that had been salvaged from a road construction site and subsequently stored for 2 years (Chiquoine et al. 2016). Cole et al. (2010) transplanted soil cores with intact bryophyte crusts in the Mojave Desert, USA. The cover and density of the bryophytes declined after transplantation, but at rates similar to the parent population, suggesting that annual declines are natural even in intact populations. In most of the aforementioned cases, inoculation hastened recovery of BSC organisms, particularly in controlled laboratory settings, with some recovery also in field studies. However, while the results were promising, the destruction of BSCs in one area to provide inoculants for another area is counterproductive in the context of large-scale arid land reclamation. Use of salvaged crusts from construction sites is promising for limited areas (Chiquoine et al. 2016). Providing sufficient supplemental water for successful large-scale reclamation in arid environments has been difficult, costly, and only minimally feasible.

Related research has investigated the potential for ex situ laboratory-grown BSC amendments for use in inoculating disturbed areas (Zhao et al. 2016; Bowker and Antoninka 2016). For example, Buttars et al. (1998) incorporated laboratory-grown cyanobacteria into alginate pellets. However, the cyanobacteria were unable to escape intact pellets. Crushing the pellets and applying them to moistened soil in the laboratory resulted in significant increases in cyanobacterial biomass, frequency, and nitrogen fixation. Incorporation of cyanobacteria into starch pellets was not successful due to poor survival during the pelletization process (Howard and Warren 1998). Kubečková et al. (2003) also grew cyanobacteria and immobilized it on hemp cloth. Laboratory trials indicated improved growth compared to alginate pellets, but in four of five field trials, there was no significant crust recovery. The general lack of success with artificial crust recovery was attributed, at least in part, to the placement of the inoculants on the soil surface where some species of cyanobacteria can be negatively affected by incident UV radiation (Castenholz and Garcia-Pichel 2012). If sensitive species occur at a depth of 1–2 mm, UV radiation is attenuated (Dor and Danin 2001). When cyanobacterial inoculants have been applied to the soil surface, rather than incorporated into the surface layer of the soil, mortality has been high. Moss protonema transplanted into the sands of the Gurbantunggut Desert of China from laboratory-grown mosses has seen some success when supplemented with liquid growth media (Xu et al. 2008). Mosses have been successfully propagated in the laboratory with frequent watering and fertilization (Antoninka et al. 2016), although field trials have not been conducted. The addition of laboratory-grown cyanobacteria to polyvinyl alcohol and a liquid soil tackifier appeared to accelerate the formation of a biocrust in a laboratory setting (Park et al. 2017).

Although some degree of success has been noted, large-scale field trials have not been attempted, and successful ex situ growth is not ubiquitous across BSC components. While we, too, have been invested in the challenge, the level of success to date, the amount of water required, and the per-acre costs have caused us to wonder whether these approaches merit further consideration in arid areas, except in critical situations where cost is not a constraint. One experience with artificial restoration highlights our concerns and lends credence to our doubts. While working in the Great Basin west of Salt Lake City, Utah, using fragments of hemp cloth embedded with dried cyanobacteria, we treated plots that had been burned by a wildfire. After one season, the chlorophyll content of the soil had increased to a level that exceeded untreated plots, suggesting success. We began to prepare a manuscript to announce the success. In the process, we decided to evaluate the plots after a second year to further document progress. However, after the second year, the untreated plots had improved to a state equal to the treated plots. Confused, we began to try to determine potential reasons. In the process, we recalled that the site had experienced frequent dust devils during the intervening period. That led us to conclude that natural passive restoration had likely occurred.

3 Passive restoration

The fact that BSCs are found in almost all environments, ranging from mesic to hyper-arid, and from temperate to extremely hot or cold, justifies the question as to how crust organisms became so spatially and climatically dispersed in the first place, and if the same processes are still operating. In general, as post-disturbance succession takes place, the initial colonizers tend to be large filamentous cyanobacteria (Belnap and Eldridge 2001). As the surface becomes stabilized, next to appear are the smaller cyanobacteria and green algae. They are often followed by small lichens propagated by vegetative diaspores. Where climatic conditions permit, larger lichens and mosses appear in later successional communities (Dümig et al. 2013). The distribution and successful establishment of these organisms is governed both by historical and contemporary factors (Leavitt and Lumbsch 2016).

Estimates of the time required for natural recovery of BSCs following disturbance have varied widely depending on the nature, periodicity, extent, and spatial and temporal distribution of the disturbance, and soil and climatic conditions. Dojani et al. (2011) reported significant recovery to a level beyond the pre-disturbance condition within 1 year (one moist season) on the Succulent Karoo semi-desert of South Africa where the upper 10 mm of the soil surface was removed. Five years following one-time human trampling, Cole (1990) noted a nearly complete recovery of visible BSC cover, although the complex pinnacled surface micro-topography attributable to many crusts had not recovered to pre-disturbance levels. Read et al. (2011) labeled as ‘surprisingly fast’ the recovery of biological soil crusts following livestock removal from an area that had been previously heavily disturbed by livestock grazing in Australia. Anderson et al. (1982) estimated that 14–18 years were adequate for recovery of a BSC following exclusion of livestock grazing in the cool Great Basin Desert, USA. In contrast, there was little evidence of recovery during the first 10 years following cessation of grazing at another Great Basin Desert location (Jeffries and Klopatek 1987). Recovery lagged 20 years following burning of a shrub community in the transition zone between the Great Basin and Mojave Deserts in southwestern Utah, USA (Callison et al. 1985). Belnap (1993) estimated that full recovery of BSCs in the Great Basin Desert, including visual as well as functional characteristics, could require as long as 30–40 years for the cyanobacterial component, 45–85 years for lichens, and 250 years for mosses. Fifty-six years following abandonment of a military training camp in the Sonoran Desert, USA, a cyanobacterial crust had not recovered to levels typical of adjacent undisturbed areas (Kade and Warren 2002). In the Mojave Desert, USA, according to measurements taken inside and outside of tank tracks created during training for World War II, and assuming a worst-case linear trajectory scenario, full recovery of the cyanobacterial component of the BSC was estimated to require up to 85–120 years (Belnap and Warren 2002).

Similar temporal patterns of BSC recovery following disturbance have been recorded in other regions. In Australia, near complete recovery was documented after 20 years on pastures that had been grazed moderately, while heavily grazed pastures recovered at a much slower rate (Read et al. 2011). Eldridge and Ferris (1999) suggested that at least 60 years would be required for full recovery of lichens at a nuclear test site in the Great Victoria Desert of Australia. In an extreme case, Lalley and Viles (2008) estimated that full recovery of lichens in badly disturbed truck ruts in the hyper-arid Namib Desert could take up to 530 years without climatic or anthropogenic intervention. It is important to note, however, that the rate of recovery is likely dependent on the arrival of viable propagules onto suitable substrates at times consistent with adequate moisture. Such conditions may occur only infrequently in the drier and hotter arid zones. We have personally witnessed significant recovery of crust organisms within 2 years following wildfire in the Great Basin Desert, USA, when suitable conditions prevailed.

Regardless of the timeframe required, recovery is dependent on several factors: (1) arrival of suitable propagules, (2) existence of an appropriate substrate on which to establish, including soil texture and chemistry, and (3) timing of the arrival of propagules in relation to cyclical soil moisture conditions suitable for establishment. The failure of any one of the necessary components may substantially delay successful re-establishment.

4 Aerobiology

As early as 1846, Charles Darwin collected dust from surfaces of HMS Beagle during one of his voyages of exploration, and reported 17 different organisms (Darwin 1846). Almost a century and a half following Darwin’s original discovery, on an ocean voyage from the Polish Antarctic Station on King George Island, South Shetland Island to Gdynia, Poland, scientists intended to collect pollen grains daily on blotter paper. In addition to pollen, however, the scientists commonly found abundant non-lichenized fungal spores, lichen thallus fragments, isidia, and soredia in all samples, indicating widespread distribution of airborne lichen propagules very long distances from terrestrial environments (Harmata and Olech 1991). Meier and Lindbergh (1935) collected airborne organisms from a fixed-wing aircraft on a flight over the Arctic from Maine to Denmark. Near the same time as the Polish discovery, the field of aerobiology was established, originally emphasizing studies of airborne fungi, bacteria, and viruses associated with respiratory diseases from indoor environments (Benninghoff 1991).

Subsequently, the field began to evaluate other potential airborne allergens including protozoans, minute arthropods, algae, and cyanobacteria in the atmosphere, and began to evaluate the seasonality and other factors affecting their presence. As a consequence, the presence of large numbers of cyanobacteria and algae have been documented as present in indoor and outdoor airborne environments ranging from low to high altitudes above the Earth (Schlichting 1969; Lee and Eggleston 1989; Sharma et al. 2007; Genitsaris et al. 2011; Després et al. 2012; Tesson et al. 2016; Lewandowska et al. 2017). Recent studies have reported the presence of hundreds of BSC taxa and thousands of individual organisms in dust samples collected from the external surfaces of homes around the USA (Barberán et al. 2015). It has been recently suggested that some organisms may go through multiple generations while in the atmosphere, such that the atmosphere becomes a truly aerial habitat (Womack et al. 2010).

Organisms that achieve airborne status may come from a variety of potential sources including plant surfaces, animals, water, and soil (Pearce et al. 2009). Unsurprisingly, many of the species documented in the atmosphere are also common in BSC communities. Airborne BSC organisms suspended by wind storms or other factors may be transported and deposited almost anywhere. Bioaerosols have been reported as being transported by dust storms in the western USA (Hallar et al. 2011). Fungi may be transported long distance by wind (Marshall 1997; Golan and Pringle 2017). Algae and cyanobacteria have been reported to occur on high latitude and high elevation glaciers from the Arctic to the Antarctic (Marshall and Chalmers 1997; Harding et al. 2011; Kvíderová 2012; Takeuchi 2013; Vonnahme et al. 2016). They have also been collected from building facades (Samad and Adhikary 2008; Sethi et al. 2012), stone monuments and statues (Lamenti et al. 2000; Tomaselli et al. 2000; Macedo et al. 2009; Verma et al. 2014), cave and catacomb walls (Castellani 2005), exposed rocks (Danin 1999), and plant surfaces (Sethi et al. 2012; McGorum et al. 2015). In addition to algae and cyanobacteria, other BSC components can also be dispersed by wind. These include asexual reproductive lichen fragments, isidia, soredia, and/or lichen-forming fungal spores (Bailey 1966; Büdel and Wessels 1986; Heinken 1999; Tormo et al. 2001; Bannister and Blanchon 2003; Hugonnot and Celle 2012; Leavitt and Lumbsch 2016). In addition, spores, gametophyte fragments, and specialized asexual diaspores of bryophytes (Marshall and Convey 1997; Stark 2003; Laaka-Lindberg et al. 2003; Pohjamo et al. 2006; Lönnell et al. 2012) have been reported. Airborne bacteria and fungi have even been found in the atmosphere of salt mines 400 m below the surface (Gębarowska et al. 2017). The pattern of airborne dispersal of BSC propagules has resulted in many species occurring in both the northern and southern Polar regions in Iceland, and extreme southern Chile (Piñeiro et al. 2012).

5 Atmospheric mixing and dispersion

A logical question may arise as to how BSC organisms are able to achieve airborne status. Many people living in arid and hyper-arid regions of the world have, at one time or another, heard stories of, or personally witnessed, dust storms that develop when strong non-convective horizontal winds blowing over unconsolidated or even intact soil surfaces pick up large quantities of soil (Fig. 1b; Kok et al. 2012). Although not at all limited to the Dust Bowl era, such conditions prevailed in the 1930’s in the North America (McLeman et al. 2014). Similarly, strong dust storms have been recorded in Alaska (Nickling 1978), China (Qian et al. 2002; Wang et al. 2004), Australia (McTainsh et al. 1998; Ekström et al. 2004), Africa (Prospero and Mayor-Bracero 2013), and the Middle East (Orlovsky et al. 2005; Almuhanna 2015; Sissakian et al. 2013). On a smaller, but much more common scale, dust may be lifted into the atmosphere by strong vertical vortices or ‘dust devils’ (Fig. 1c) (Metzger et al. 2011; Kok et al. 2012; Horton et al. 2016). The maximum wind speeds of dust devils have been recorded as 22 ms−1 (45 mph) (Schwiesow and Cupp 1976).

Once airborne (Fig. 1d), dust particles and the BSC organisms that accompany them are subject to a variety of forces that carry them between hemispheres, continents, and climatic zones (Griffin et al. 2002; Prospero and Lamb 2003; Kellogg and Griffin 2006; Uno et al. 2009; Pointing and Belnap 2014). Near the Earth’s surface, airborne particles are carried predominantly by trade winds, which were given their name because of the effect they had on global oceanic trade prior to the advent of fossil fuel powered transport. Trade winds exist in six major belts which circle the globe. Between the equator and 30° north or south latitude, the trade winds generally blow from east to west; between 30° and 60° latitude, the winds then shift to from west to east; between 60° north or south latitude and the respective poles, easterly winds again prevail. The major jet streams exist at about 9–15 km above the Earth’s surface and blow from west to east (Lewis 2003). They meander north or south, and may cross between the northern and southern hemispheres (Rangarajan and Eapen 2012). Other than the trade winds and jet streams, a primary force mixing the atmosphere within the northern and southern hemispheres is the global Hadley, Ferrel, and polar cells (Kjellsson and Döös 2012; Huang and McElroy 2014) which correspond latitudinally to the trade wind belts. Hadley cells begin where warm air rises near the equator, generally resulting in heavy rainfall. After reaching the upper atmosphere, the Hadley cells carry the flow of air poleward. At approximately 30° north and south latitude, the Hadley cells diverge earthward, converging with the downward flow of the Ferrel cells bringing air masses from greater latitudes. The result in both hemispheres is a very large body of descending dry air and high pressure. As descending air masses typically offer little precipitation, the zones of convergence correspond with some of the world’s most recognized arid zones. The dry air moves poleward after reaching the Earth’s surface. As the Ferrel cells pass over the Earth’s surface, they collect moisture until they reach approximately 60° north and south latitude, where the air masses ascend after converging with the polar cells. In order to continue the ascent, the air masses lose moisture, and precipitation increases. The polar cells descend earthward near the poles, a region also widely known for aridity.

Given the forces mixing the atmosphere, and the likelihood for BSC propagules to be airborne, there can be little doubt that organisms originating from almost any given location have the potential to be deposited anywhere on Earth (Jungblut et al. 2010; Barberán et al. 2014; Herbold et al. 2014). Carson and Brown (1976) found little correlation between the diversity of airborne algae, and soil algae at corresponding altitudes on the Island of Hawaii, suggesting atmospheric mixing of airborne organisms. Evidence of mixing can also be seen on a global scale by the bipolar similarity of BSC species in the Arctic and Antarctic (Galloway and Aptroot 1995; Søchting and Olech 1995; Jungblut et al. 2012; Fernández-Mendoza and Pritzen 2013). Dust deposited in Antarctica originates in Patagonia, Australia, and the Northern Hemisphere (Li et al. 2008). Dust originating during dust storms in China and the Middle East has been documented as arriving in Japan within just a few days (Lee et al. 2006). Dust from the Middle East has been recorded in the Caribbean (Doherty et al. 2008; Swarf et al. 2014) and the southeastern USA (Prospero 1999). Many BSC propagules carried with dust can survive long periods of desiccation (Potts 1994; Holzinger and Karsten 2013; Rajeev et al. 2013), thus becoming immigrants deposited in distant BSC communities globally (Rosselli et al. 2015; Rahav et al. 2016). For example, lichen species of South African origin are now present in Australia and South America (Amo de Paz et al. 2012). Similarity of BSC communities is better predicted by the so-called dust highways than by the proximity of potential source species (Muñoz et al. 2004). Dust and microbial deposition are both seasonal (Sharma et al. 2006a, b; Dubey et al. 2010; Sahu and Tangutur 2015) and cyclical over time (Rousseau et al. 2007).

The apparent airborne and global distribution of BSC propagules should not be construed to imply that BSC species composition will be the same worldwide, nor that natural recovery of BSC’s will be necessarily rapid. The distribution of BSC propagules is shaped by the dynamic interplay of a range of factors operating across multiple temporal scales. That many propagules are distributed globally is apparently true. However, whether they will develop and thrive in a new location is still dependent on being deposited on a suitable substrate determined by chemistry, texture, fertility, particle and pore size analysis, moisture content and seasonality, and temperature, among other factors that determine the suitability of habitat for the BSC taxa in question (Miller and McDaniel 2004; Nagy et al. 2005; Kharkongor and Ramanujam 2014; Johansson et al. 2015). For example, it is hardly realistic to expect most BSC species adapted to the frigid conditions of polar regions to survive and persist in hot deserts, and vice versa.

6 The nexus of aerobiology and land reclamation

The use of corn stalk fences and wheat straw checkerboard sand barriers to stabilize dune sands have been successfully used for years in China (Qiu et al. 2004; Zhang et al. 2004; Li et al. 2006). These barriers create turbulence in the flow of wind across the dune surfaces, and cause the deposition of sand and soil, and associated BSC organisms. Researchers have discovered that biological soil crust organisms precipitated in this fashion can successfully colonize stabilized dunes (Li et al. 2003; Guo et al. 2008; Zhang et al. 2014). Li et al. (2016) reported the results of projects in a severely disturbed sandy habitat of the Qinghai–Tibet Plateau near the Yellow River in China. Grass and woody species were planted to increase wind turbulence and promote the deposition of eolian sediments. Within as little as 10 years, the clay and humus content of the soil had increased by 15%, and a BSC community, comprised primarily of lichens, mosses, and algae, had become established.

One must bear in mind that while airborne BSC propagules may provide an answer to the restoration of BSCs in many situations, their presence and composition depends on climatic conditions in locations that may be very far away. As discussed, BSC propagules may originate from distant continents and hemispheres. BSC organisms from a specific soil type, chemistry, and alkalinity may not always be suitable for other locations. The arrival of appropriate propagules is likely episodic, seasonal, and less frequent than desired (Favero-Longo et al. 2014; Gosselin et al. 2016; Weber et al. 2016a). However, there can be little doubt that airborne propagules are found in the atmosphere circling the globe. Whether they will be deposited in sufficient quantities, in the right species composition, and at the right season for any specific area remains unknown.

7 Conclusions

Over the years, millions of dollars have been expended on various approaches to soil crust establishment and restoration, often culminating in the production and application of large quantities of inoculants. Many of these approaches have been less than successful to one degree or another. Those approaches that did not fail outright have been so dependent on large quantities of water for production, application, and maintenance that they have not been practical for broad-scale application to arid and semiarid environments. Several approaches to restoration have depended on the destruction of the soil crust community in one area in order to restore another. A review of the field of aerobiology seems to indicate that there may be good reasons for considering and incorporating passive restoration more seriously as a viable method to support BSC recovery. Propagules of many, if not most BSC organisms are already present and circulating the globe in the atmosphere. Perhaps, now we should carefully examine the possibilities of a more natural approach to crust restoration. Whether we artificially produce and apply inoculants, or rely on natural, passive dispersal, the overall success depends on coordination of inoculation with appropriate environmental conditions. At any given location, regardless of the mode of inoculation, success depends on receiving adequate moisture at the right time of year, the availability of appropriate substrata, a host of other environmental factors, and some measure of better controlling anthropogenic disturbance to BSC communities. We anticipate that incorporating principles of aerobiology and passive dispersal into the BSC restoration paradigm will facilitate more effective and less costly management and recovery of BSCs.