C, N, and P Nutrient Cycling in Drylands

Abstract

Drylands play a significant role in the global biogeochemical cycling of nutrients (carbon, nitrogen, and phosphorus) through abiotic (geological, atmospheric, and hydrological) and biotic (animals, insects, plants, and microorganisms) pathways. They act as important carbon reservoirs and are estimated to store over 30% of the global soil organic carbon reserve. However, nitrogen and phosphorus availability are major limiting factors for biological activity in these oligotrophic environments, affecting community structure, species diversity, and other ecosystem functions (e.g., nutrient cycling and productivity). Nutrient cycling in desert soils is primarily achieved by plant and microbial communities, in particular soil microbial communities, biological soil crusts, hypoliths, and endoliths. Drylands are highly sensitive and prone to disturbance and land degradation resulting from desertification. Changes induced by climate (e.g., precipitation and temperature), structural and temporal variability (nutrient accumulation and distribution of minerals, seasonal variation, and differences in turnover rates), and human activity often alter nutrient cycles that negatively affect the structure and function of these ecosystems (e.g., decreasing carbon storage capacity, increasing NOx emissions, and reducing phosphorus cycling). Comprehending the extent, nature, magnitude, and reversibility of such changes is urgent, given the global importance of drylands in terms of carbon sequestration, greenhouse gas emissions, ecology and biodiversity, and human habitation.

The focus of this chapter is motivated by the high sensitivity of drylands to perturbations affecting the past and current nutrient cycling of most of these ecosystems around the world. It shows how changes in the abiotic and mostly biotic components and their interactions of each particular dryland’s microclime ecosystems can affect the dynamics of their nutrient cycling both in time and space. Thus, in this chapter, we present a general discussion of the dynamics of the soil C, N, and P cycles in drylands with special attention to the role of Biological Soil Crusts, the influence of hydration-desiccation pulses, and the impact of Climate Change on nutrient cycling.

7.1 Carbon in Drylands Soils

Studies have indicated that drylands currently act as a large carbon (C) sink and play a more significant role in the terrestrial C balance than previously expected (Lal 2001; Wohlfahrt et al. 2008; Rotenberg and Yakir 2010). Dryland ecosystems are an important component of the global C cycle as they store ~32% (~470 Pg; Plaza et al. 2018a) of the global soil organic carbon (SOC), account for about one-third of the global vegetation C storage (Allen-Diaz et al. 1996), and represent 30–35% of terrestrial net primary production (NPP; Field et al. 1998). Drylands, therefore, dominate the positive global land CO₂ sink, although large fractions of the interannual variability (IAV) of net CO₂ flux have been mostly associated with variations in gross primary production (GPP; Beer et al. 2010; Yao et al. 2020). Model projects showed that the majority of dryland GPP variability is attributed to precipitation and air temperature (Ahlström et al. 2015; Zhang et al. 2016a; Yao et al. 2020), although other environmental variables such as changes in ecosystem types also play relatively small roles in regulating GPP variability (Yao et al. 2020). With the increasing frequency of extreme climatic events, the IAV of GPP is also projected to increase and will likely cause significant impacts on the global C cycle (Zscheischler et al. 2014). Although the physical-chemical properties of dryland soils limit their potential to store C (Ewing et al. 2006; Serrano-Ortiz et al. 2012; Weil and Brady 2017; Plaza et al. 2018a), their large coverage of the global land area, and the fact that many of these soils have been degraded, means that drylands have the greatest potential to sequester C (Scurlock and Hall 1998; Rosenberg et al. 1999). In fact, the amount of SOC present in drylands is ~42 times more than what is added into the atmosphere through anthropogenic activities, estimated at 11.3 Pg C year^–1 in 2017 (Lal 2004a; Le Quéré et al. 2018). Studies have indicated that drylands currently act as a large C sink and play a more significant role in the terrestrial C balance than previously expected (Lal 2001; Wohlfahrt et al. 2008; Rotenberg and Yakir 2010).

In this chapter, we present a general discussion of the dynamics of the soil C cycle in drylands with special attention to C sequestration and climate change in arid environments. Although greenhouse gases methane (CH₄) and nitrous oxide (N₂O) are important components in the C cycle, carbon dioxide (CO₂) is the most prevalent greenhouse gas in the atmosphere and will be the primary gas considered when discussing C sequestration.

Soil C pools in arid environments exist as three main components: (1) soil organic C (SOC); (2) soil inorganic C (SIC); and (3) biomass organic C (Serrano-Ortiz et al. 2012). Important features and current stocks will be briefly discussed for each pool in the following sections.

7.1.1 Soil Organic Carbon in Drylands Soils

SOC is a strong determinant of soil quality, particularly in arid environments, as it influences the physical, chemical, and biological properties of soil (Gaitán et al. 2019) and is critical for improving soil fertility (Zhang and Shao 2014). As the SOC pool is strongly affected by precipitation and temperature (Lal 2002), SOC storage tends to decrease exponentially with an increase in temperature, but increases with increments in soil water content (Lal 2002; Follett et al. 2012; Yan et al. 2017; Chatterjee et al. 2019). Consequently, dryland soils contain small amounts of SOC, often less than 0.5% of the soil mass resulting in typical densities of 0–15 kg m^–2 (Lal 2002, 2004a). However, due to the vast extension of dryland ecosystems soils are estimated to contain ~ 646 ± 9 Pg of SOC to a 2-m depth, representing about 32% of the world’s total SOC pool (Plaza et al. 2018a; Lal 2019).

Arid and hyper-arid soils contain ~113 Pg organic C, whereas semiarid and sub-humid areas store ~ 318 Pg SOC (Plaza et al. 2018b). Studies have demonstrated that the SOC density is lower for bare soils (1–3 kg C m^–2; Rasmussen 2006; Woomer et al. 2004) compared to shrublands (2–6 kg C m^–2; Chen et al. 2007; Wiesmeier et al. 2011) and grasslands (5 ± 2 kg C m^–2; Chen et al. 2007; He et al. 2008).

7.1.2 Soil Inorganic C in Drylands Soils

SIC is also a critical component of arid regions and plays an important role in carbon sequestration and climate alleviation (Lal 2004b). Dryland SIC pools are estimated to be higher (1237 ± 15 Pg; Plaza et al. 2018a) than SOC pools and may even exceed SOC by a factor of 10 in some arid areas (Schlesinger 1985, 2006; Scharpenseel et al. 2000). The SIC stocks are generally highest for arid (487 ± 7 Pg) and semiarid (456 ± 7 Pg) environments (Plaza et al. 2018a) where the SIC pool is mainly present as lithogenic inorganic carbonates (LIC) that originates as detritus from parent materials (i.e., limestone), and pedogenic inorganic carbonates (PIC) formed by dissolution and re-precipitation of LIC or by the dissolution of CO₂ into bicarbonate (HCO₃^–) followed by precipitation with Ca²⁺ and/or Mg²⁺ originating from non-LIC minerals (e.g., silicate weathering, dust, and fertilizers) (Marion 1989; Schlesinger 1985, 2002; Tan et al. 2014; Wang et al. 2015; Sahrawat 2003; An et al. 2019).

These carbonates tend to precipitate at relatively shallow depths as a result of low precipitation and poor leaching of soils (Gocke et al. 2011). Consequently, about 80% of the global SIC pool is found in drylands (Eswaran et al. 2000; Plaza et al. 2018a). Reported SIC stocks can be modified significantly taking into account anthropogenic practices (Serrano-Ortiz et al. 2012). For example, the SIC pool can be affected by land management practices such as afforestation, irrigation, fertilization, and liming (Sanderman 2012; Gao et al. 2017). Such activities can cause elevated levels of CO₂ in soils, resulting in significant formation and precipitation of secondary carbonates, thereby contributing to soil C sequestration (Denef et al. 2008; Sanderman 2012; Gao et al. 2017).

7.1.3 Biomass Organic C in Dryland Soils

7.1.3.1 Vegetation

Dryland vegetation is highly variable and can range from barren or sparsely vegetated desert to grasslands, shrublands and savannahs, croplands, and dry woodlands (FAO 2004; Lal 2004a; Mander et al. 2017). Vegetation cover in drylands is influenced by various factors including drought stress, variations in annual temperatures, and precipitation intensity and frequency (FAO 2004; Lal 2004a; Mander et al. 2017). These factors strongly impact above-ground biomass productivity and therefore the level of SOC in soil.

Carbon storage for dryland vegetation is overall low (Eswaran et al. 2000), with the global average and maximum stock estimated at ~65 and 81 Pg C, respectively (Safriel et al. 2005; Serrano-Ortiz et al. 2012). However, these values can be modified significantly when considering anthropogenic activities, like grazing and desertification, which may reduce the biomass C pool by 10–20% of the given value (Serrano-Ortiz et al. 2012; An et al. 2019). Biomass C density can differ between arid regions, depending on vegetation coverage and land use. For example, hyper-arid and arid deserts have the capacity to store 0.04–0.40 kg C m^–2 in biomass (Woomer et al. 2004; Fan et al. 2008; Perez-Quezada et al. 2011), while grasslands and sub-humid forested drylands have a biomass C density of ~1 (He et al. 2008) and 4–5 kg C m^–2 (Glenday 2008), respectively.

7.1.3.2 Microbial Autotrophs

As most drylands represent sparse vegetation cover, autotrophic microbes (soil and BSC) act as important primary producers (Liu et al. 2018). Autotrophic CO₂ fixation processes are significant for C accumulation in dryland soils that may influence ecosystem succession processes (Agarwal et al. 2017; Liu et al. 2018). In fact, microbial autotrophy accounts for ~ 4% of total C sequestered by terrestrial ecosystems per year, of which drylands comprise a substantial fraction (Yuan et al. 2012). Microbial autotrophs are distributed amongst both bacteria and archaea with highly diverse phylogeny, metabolic activities, and ecological variants (Hügler and Sievert 2011). For example, studies have shown that autotrophic microbes in drylands belong to the taxa Actinobacteria, Proteobacteria, Chloroflexi, Acidobacteria, Gemmatimonadetes, Firmicutes, Thaumarchaeota, Nitrospirae, Planctomycetes, and Bacteroidetes (Yang et al. 2017; Liu et al. 2018).

Even though the Calvin cycle is the predominant pathway utilized by microbes in nutrient-rich conditions, many autotrophic microbes in drylands fix C through the energy-conserving reductive citrate cycle, reductive acetyl-CoA cycle and 3-hydroxypropionate bi-cycle pathway (Agarwal et al. 2017; Liu et al. 2018). Liu et al. (2018) were able to demonstrate the capacity of desert microbial autotrophs to directly incorporate ¹³CO₂ into SOC using the three above-mentioned pathways, which accounted for ~4% of the atmospheric CO₂ absorbed by desert soil. In addition, the study showed that the efficiency of autotrophic CO₂ fixation was impacted by soil properties, the autotrophic composition, and the abundance of genes associated with the CO₂ fixation pathways. Their results highlighted the underestimated importance of microbial autotrophy in the C cycle and storage in drylands.

7.1.4 Carbon Sequestration and Loss: The Impact of Abiotic and Biotic Factors

Carbon sequestration in drylands involves the transfer of atmospheric CO₂ into both SOC and SIC pools via management of vegetation, soil, and water resources that lead to a positive C budget (Lal 2009). Mechanisms involved in SOC and SIC sequestration, and the rate of each process are discussed by Lal (2009). In general, primary producers/plants convert atmospheric CO₂ to complex organic molecules through photosynthesis; these molecules enter the soil C cycle as decaying organic matter (OM; Batlle-Aguilar et al. 2011; Serrano-Ortiz et al. 2012). A significant portion of the OM is directly used to sustain energy for pedo-and microfauna metabolism and is released back to the atmosphere as a by-product of autotrophic (plant) and heterotrophic (e.g., microbial) respiration (Batlle-Aguilar et al. 2011; Thomey et al. 2014). Some of the soil C is assimilated by vegetation and finally transferred to the soil as plant litter, becoming part of SOM (Porporato et al. 2003). In addition to biological processes, there is growing evidence to suggest that abiotic processes contribute to CO₂ fluxes in soils and may even dominate the flux during dry seasons (Kowalski et al. 2008; Serrano-Ortiz et al. 2012; Cueva et al. 2019). The exact physical and biogeochemical mechanisms that promote CO₂ capture in soils are still uncertain and warrant future research to accurately understand and quantify the C balance, from local to global scales (Cueva et al. 2019).

Soil C stock in drylands is strongly impacted by biotic and abiotic processes, factors (e.g., climate, vegetation, soil properties), causes (e.g., urbanization, wildfires), and interactions among them (Lal 2019), that primarily control C fluxes between the atmosphere and both SOC and SIC pools via management of vegetation, soil and water resources that lead to a positive C budget (Lal 2019). The C stock is also affected by erosional processes, both aeolian and hydrologic, which influence transport, redistribution, and deposition of C over the landscape (Lal 2019).

Significant progress has been made to understand the controls on regional and global patterns of soil SOC in drylands. Correlative analyses and/or structural equation modeling (SEM) across natural environmental gradients indicate that SOC is largely controlled by climatic factors such as precipitation and temperature, as well as soil properties and plant productivity (White et al. 2009; He et al. 2014; Wang et al. 2014a–d; Mureva et al. 2018; Gaitán et al. 2019; Smith and Waring 2019; Zhu et al. 2019). For example, precipitation influences SOC storage by constraining primary productivity and decomposition (Wynn et al. 2006; Yang et al. 2007), whereas higher temperatures accelerate the microbial decomposition of SOM, thereby causing C loss (Giardina and Ryan 2000).

In addition to climate, soil properties, such as soil mineralogical characteristics (Smith and Waring 2019), pH (Min et al. 2014; Ou et al. 2017), bulk density (Feng et al. 2002), total phosphorus (Tian et al. 2017), and soil moisture content (Wang et al. 2014a–d), are strongly related to SOC. It should be noted that the main factors controlling SOC concentration may differ between regions and even within the same ecosystem (Cable et al. 2011), or can be scale dependent (Dai and Huang 2006; Liu et al. 2006; Evans et al. 2011; Wang et al. 2013; Qin et al. 2016). An important reason for such discrepancies is that soil CO₂ efflux is a combined result of biotic processes and abiotic factors (Ma et al. 2010, 2017; Gaitán et al. 2019), each of which exhibits its own flux behavior at various time scales and responds differently to the environment (Ryan and Law 2005; Li et al. 2010) to control SOC concentration and dynamics.

7.2 Nitrogen in Dryland Soils

Second to water, nitrogen (N) availability is the main limiting factor of biological activity in oligotrophic arid environments (Whitford 2002). Microorganisms in open soils, root systems, or cryptic niches (e.g., biological soil crusts, hypoliths, and endoliths) are the main providers of atmospheric fixed N but also act as the key players in other N-cycling processes such as nitrification and denitrification.

7.2.1 Biological Nitrogen Fixation (BNF) as N Input in Drylands

The main nitrogen input in many hyperarid environments comes from biological nitrogen fixation (BNF) (Johnson et al. 2005; Pointing and Belnap 2012). The BNF is an anaerobic process that is performed only by prokaryotes. The oxygen-sensitive nitrogenase complex catalyzes the reduction of atmospheric N₂ to bioavailable ammonia (NH₃). This is a highly energy-demanding process for cells (16 ATP molecules and 8 electrons per reduced N₂ molecule; Zehr et al. 2003), particularly in desert environments that are nutrient deprived. Acetylene reduction assays (ARA) are commonly used to measure the nitrogenase activity in oligotrophic environments via the reduction of acetylene to ethylene (Aranibar et al. 2003) and calculating the rates by which ¹⁵N is incorporated into cells (Mayland and McIntosh 1966; Caputa et al. 2013; Alcamán-Arias et al. 2018).

The most widely used phylogenetic marker to study N-fixation is the nifH gene as its phylogeny is fairly concordant with that of the 16S rRNA gene (Zehr et al. 2003). Nitrogen-fixing bacteria (diazotrophs) belong to several phyla including the autotrophic Cyanobacteria and Chlorobi, as well as to heterotrophic Actinobacteria, Firmicutes, and Proteobacteria (α-, γ-, ε-classes; see Hartley et al. 2007); all of which have been detected in hot and cold deserts (Cowan et al. 2011; Ronca et al. 2015; Crits-Christoph et al. 2016; Lacap-Bugler et al. 2017; Meslier et al. 2018). Also, several archaeal taxa such as Methanobacteria, Methanococci, and Methanomicrobia have the potential to fix N (Chan et al. 2013; Wei et al. 2016) as recently demonstrated in wetland soils (Bae et al. 2018).

Biological soil crusts (BSCs) are abundant in cold and hot deserts (Belnap 2001a; Pointing and Belnap 2012; Makhalanyane et al. 2015; Chap. 3) and can contribute to ~30% of biologically fixed nitrogen in terrestrial ecosystems (Elbert et al. 2012). In cyanobacterial dominated crusts N is primarily fixed by filamentous heterocystous (e.g., Anabaena, Nostoc, Scytonema) and non-heterocystous (e.g., Microcoleus, Chroococcidiopsis, Phormidium) cyanobacteria (Belnap 2002; Wei et al. 2016; Lacap-Bugler et al. 2017; Mogul et al. 2017), although other organisms such as chlorophyte algae, heterotrophic bacteria, fungi, mosses and/or lichens are also present (Belnap 2003; Pointing and Belnap 2012). Studies have shown that N-fixation in cyanobacterial–algal crusts from the Tengger and the Gurbantunggut Deserts presented the highest nitrogenase activity (up to 16.6 mmol m⁻² h⁻¹), followed by lichens (up to 6.9 mmol m⁻² h⁻¹) and mosses (up to 2.6 mmol m⁻² h⁻¹; Wu et al. 2009; Su et al. 2011). In contrast, lichen-dominated crusts in the Colorado Plateau were found to be the main N-fixers and the amount of N fixed was ten-fold higher than that of cyanobacteria-dominated crusts (Belnap 2002). Looking at ammonium oxidation (AO) in two types of BSCs from the Colorado Plateau, Johnson et al. (2005) found that N₂ fixation potential was appreciably higher (6.5–48 μmol C₂H₂ m^–2 h^–1) in dark crusts, dominated by heterocystous cyanobacteria, than those in light crusts.

N-fixation and N-fixing activity in drylands are influenced by several environmental factors, of which water availability and temperature appear to be the most important parameters (Belnap 2001b; Hartley et al. 2007). Cyanobacteria and other heterotrophic bacteria are physiologically active only when water is available, thereby controlling photosynthesis and ultimately N-fixation (Belnap 2001b; Hartley et al. 2007). Moisture levels needed to initiate and optimize N-fixation vary widely with species, habitats, and pre-existing conditions (Belnap 2001b). N-fixation rates are also limited by temperature extremes. Most nitrogenase activity occurs at –5 to 30 °C, with the optimum at 20–28 °C for the majority of drylands (Belnap 2001b; Wu et al. 2009; Schwabedissen et al. 2017). Low temperatures can reduce photosynthetic rates and thus available ATP and reductant pools, creating a lag time before N-fixation is initiated (Belnap 2001b). In contrast, at higher temperatures (i.e., above minimum temperature for species) N-fixation rates show a strong, positive response to increasing air temperature until an upper limit is reached, after which rates quickly decline (Belnap 2001b).

Other factors that affect N-fixation in drylands include pH, salinity, and nutrient content (Hartley et al. 2007). Nitrogenase activity in soil cyanobacteria and BSCs are greatest above pH 7 (Hartley et al. 2007; Schwabedissen et al. 2017); however, nitrogenase activity is reduced at high (8–10) and low pH (see Belnap 2001a for further explanations). Effects of salinity on nitrogenase activity in drylands are not well studied and show mixed results. For example, N-fixing cyanobacteria in Utah deserts preferred soils with high electrical conductivity (70 dS m^–1; Anderson et al. 1982). In contrast, experimental addition of NaCl to cyanobacterial lichen crusts in the Chihuahuan desert inhibited nitrogenase activity (Delwiche and Wijler 1956). Nutrient effects on nitrogenase activity vary in drylands and largely depend on the element. Phosphorus (P) and potassium (K) additions can stimulate cyanobacterial nitrogenase activity, likely through the stimulation of ATP synthesis (Dodds et al. 1995). Low amounts of zinc (Zn), cobalt (Co), molybdenum (Mo), and iron (Fe) can also stimulate cyanobacterial nitrogenase activity, whereas higher levels of the same elements have adverse effects (Granhall 1981; Dodds et al. 1995). Glucose (i.e., carbon) additions to soils and BSCs have been shown to increase heterotrophic nitrogenase activity, suggesting that C sources are essential in N dynamics (Hartley and Schlesinger 2002; Billings et al. 2003; Hartley et al. 2007).

Hypolithic microbial communities (biofilms on quartz and translucent rocks) represent major contributions to organic biomass in extreme deserts where communities are dominated by cyanobacterial photoautotrophs (Chan et al. 2012), although red Chloroflexi-dominated hypoliths have been encountered in the Atacama Desert (Lacap et al. 2011). In the Mojave Desert, hypolithic dinitrogen fixation potential was demonstrated using ARA measurements (Schlesinger et al. 2003), while a recent shotgun metagenome study revealed that Namib Desert hypolithic communities could mediate the full N-cycle (apart from ANAMMOX; Vikram et al. 2016). These results were later supported by stable N isotope (δ¹⁵N) measurements of Namib Desert hypolithic biomass, and surface and subsurface soils over a 3 year period across dune and gravel plain biotopes (Ramond et al. 2018). The authors demonstrated that photoautotrophic hypolithic communities are the main contributors to N₂ fixation for plant productivity events in the Namib Desert. In accordance with these findings, metatranscriptome profiling of Namib Desert soils revealed an active microbial community dominated by non-photosynthetic bacteria where nitrate appeared to be the source of bioavailable N (Leon-Sobrino et al. 2019).

7.2.2 Atmospheric N Deposition and N Discharges as N Inputs in Drylands

Although BNF is the most studied and described process to generate N in drylands, other inputs such as atmospheric depositions and accumulations, environmental runoff, discharge of N-rich untreated wastewater, and/or agriculture processes are also important N sources.

Drylands act as large reservoirs for nitrate-rich salts. N is delivered to desert landscapes by either wet or dry atmospheric deposition of particles produced by gas to particle conversion, and by biological N₂ fixation (Böhlke et al. 1997; Michalski et al. 2004; Graham et al. 2008). Wet deposition of N (e.g., precipitation, sea spray, and fog) has been implicated as sources of NH₄⁺, NO₃^– and other salts in the Turpan-Hami area, northwestern China (Qin et al. 2012), and the Atacama (Michalski et al. 2004; Wang et al. 2014c, d), Mojave (Michalski et al. 2004), and Chihuahuan Deserts (Báez et al. 2007). Additionally, drylands receive external N inputs via the discharge of untreated, ammonia-rich wastewater (e.g., Negev and Tengger Deserts), or through agricultural processes where poultry manure is used for soil amendments (Abed et al. 2010; Posmanik et al. 2017; Zuo et al. 2018). Large N inputs have been deposited in areas surrounding the Gurbantunggut Desert following farming activities (Li et al. 2012a). All these practices induce nitrification and denitrification with the subsequent release of greenhouse gasses (N₂O).

7.2.3 Nitrogen Losses in Drylands

Biological fixed N does not accumulate over time and is lost to the atmosphere from drylands through several mechanisms including ammonia volatilization, wind erosion, leaching, nitrification, and denitrification (Peterjohn and Schlesinger 1990; Walvoord et al. 2003; McCalley and Sparks 2009; Wang et al. 2014a, b; Jin et al. 2015). Volatilization is considered to be the primary cause of N loss in deserts, with wind erosion contributing to 6.6% of total N loss (Peterjohn and Schlesinger 1990; McCalley and Sparks 2009). Volatile losses of N include nitric oxide (NO), nitrous oxide (N₂O), ammonia (NH₃), and dinitrogen (N₂) gas (Peterjohn and Schlesinger 1990). Of these compounds, NO, N₂O, and NH₃ are of particular consequence to the atmosphere affecting ozone levels and contributing to global warming (Logan et al. 1981; Peterjohn and Schlesinger 1990; van Amstel and Swart 1994). N loss from dryland soils through leaching was previously considered to be negligible, however, several studies challenged this assumption and demonstrated significant NH₃ leaching and accumulation in subsoil zones (Walvoord et al. 2003; Jin et al. 2015).

Nitrification and denitrification processes, especially in BSCs, are considered major contributors to N loss in dryland soils (Austin et al. 2004). BSCs have been shown to emit NO, N₂O, and nitrous acid (HONO) with amounts dependent on N₂ fixation rates, crust type, and season (Barger et al. 2005; Abed et al. 2013; Weber et al. 2015; Maier et al. 2018). Weber et al. (2015) showed that dark cyanobacteria-dominated crusts exhibit the highest NO and HONO emission fluxes and can even exceed those of neighboring uncrusted soils by 20 times. Using laboratory, field, and satellite measurement data, the authors estimated a global emission of reactive N from BSCs at ~1.7 Tg year^–1 (1.1 Tg year^–1 of NO-N and 0.6 Tg year^–1 of HONO-N), corresponding to ∼20% of global NO_x emissions from soils under natural vegetation. In the desert of northern Oman (Arabian Desert), N₂O gas was emitted from incomplete denitrification at very high potential rates of 387 ± 143 and 31 ± 6 μmol N m^–2 h^–1 from cyanobacterial and lichen crusts, respectively (Abed et al. 2013). This contributed up to 54–66% of the total produced gases during denitrification in both types of crusts. Although lower N₂O emissions have been detected from other drylands such as the Sonoran Desert (Guilbault and Matthias 1998), Great Basin (Mummey et al. 1994), and southeastern Utah (Belnap 2001b), their environmental implications should not be underestimated.

The potential for N loss from drylands is primarily affected by precipitation, substrate availability, and high soil-surface temperatures (McCalley and Sparks 2008, 2009; Yahdjian and Sala 2010). In the Mojave Desert, NH₃ dominated reactive N gas emissions were shown to be higher during the summer season, while a fall precipitation event yielded even larger N fluxes (McCalley and Sparks 2008). Further laboratory manipulations of environmental conditions on N gas emissions showed a large transient NH₃ pulse (~70–100 ng N m⁻² s⁻¹) following water addition, presumably driven by an increase in soil NH₄ concentrations. As such, NO production was boosted with maximum NO flux rates of 34 ng N m⁻² s⁻¹. Similar results have been observed for rainfall manipulated experiments in the Chihuahuan Desert (Hartley and Schlesinger 2000) and Patagonian steppe (Yahdjian and Sala 2010). N₂O flux in drylands is relatively constant over the dry seasons; however, N₂O emissions have been shown to considerably increase following precipitation and/or irrigation events (Hall et al. 2008; Chen et al. 2013; Liu et al. 2014; Yue et al. 2020). These brief “wetting-pulse” patterns of N₂O fluxes can account for the majority of annual N₂O emissions, making dryland N₂O patterns unique compared with other terrestrial ecosystems (Hu et al. 2017). Interestingly, studies have found that N gas fluxes are responsive to additions of labile C and elevated CO₂ (Schaeffer and Evans 2005; McCalley et al. 2011). Increased availability of C alters soil N dynamics by increasing biological (plant and microbial) activity and immobilization of N, thereby reducing N availability for nitrification and denitrification, and NH₃ volatilization (Schaeffer and Evans 2005; McCalley et al. 2011).

7.2.4 Nitrification and Denitrification in Desert Soils

Nitrification is a biological two-step process in which NH₃ is oxidized to nitrite (NO₂⁻) followed by the oxidation of NO₂⁻ to NO₃^–. The transformation of ammonia to NO₂⁻ is usually the rate-limiting step of the process (Kowalchuk and Stephen 2001) performed by ammonia-oxidizing bacteria (AOB) and/or archaea (AOA) (reviewed by Offre et al. 2013; Stein 2019). Subsequent oxidation of NO₂⁻ to NO₃^– is performed by nitrite-oxidizing bacteria (NOB), mostly from the genera Nitrobacter and Nitrospira. Both steps produce energy coupled to ATP synthesis. Ammonia- and nitrite oxidation may also occur simultaneously in Comamox bacteria (Daims et al. 2015; van Kessel et al. 2015).

Nitrification potential for drylands has been demonstrated in both soils and BSCs across several deserts and drylands (e.g., Colorado Plateau, Great Basin, and Negev and Sonoran Deserts; Nejidat 2005; Strauss et al. 2012; Marusenko et al. 2013, 2015; Delgado-Baquerizo et al. 2016). Interestingly, results suggest that temperature and aridity are important modulators of AO communities and that AOA abundance dominates over that of AOB in bare soils at more extreme conditions (i.e., higher temperature and aridity; Marusenko et al. 2013; Delgado-Baquerizo et al. 2016). Conversely, vegetated dryland soils show a higher abundance and richness of AOB than AOA due to higher OM and available N contents (Delgado-Baquerizo et al. 2016). Collectively, results suggest that niche differentiation plays a key role between AOA and AOB communities (Marusenko et al. 2013, 2015; Zhou et al. 2016; Delgado-Baquerizo et al. 2016). Furthermore, AO estimates (mean of 110 μmol N m^–2 h^–1) in BSCs from the Colorado Plateau and the Mojave, Sonoran, and Chihuahuan Deserts indicate that the magnitude of N cycling processes (e.g., nitrification) is subject to seasonal changes where soil temperature and moisture levels affect microbial activity (Strauss et al. 2012). Crust type and oxygen availability have also been implicated as important factors in ammonia oxidation fluxes (Johnson et al. 2005). For example, Johnson et al. (2005) found that AO rates (measured at depth of 2–3 mm) from light cyanobacterial crusts (53.4 ± 28.1 mmol N m^–2 h^–1 in light and 28.6 ± 13.0 mmol N m^–2 h^–1 in dark) were higher than for dark cyanobacterial crusts (42.0 ± 21.1 mmol N m^–2 h^–1 in light and 14.8 ± 10.2 mmol N m^–2 h^–1 in dark), irrespective of the incubation strategy (Johnson et al. 2005). It should be noted that potential rates were based on wet crusts and are likely to be overestimated.

Nitrate is removed from the immediate environment by processes such as denitrification, assimilation, or anaerobic ammonia oxidation (anammox). Denitrification, the biological process in which NO₂^–, NO, and N₂O are successively reduced to N₂ gas, is a facultative anaerobic process performed by heterotrophic bacteria (Williams et al. 1992). This process is highly dependent on NO₃^– and C source availability (Williams et al. 1992). In addition to bacteria, several fungi are able to perform denitrification during anaerobic respiration (Shoun et al. 2012). Drylands are not considered as denitrification hotspots, although denitrification potential and denitrifying enzyme activity have been detected in grasslands (Santa Barbara, CA; Homyak et al. 2016), semiarid (“El Romeral,” Cajón del Maipo, Chile; Orlando et al. 2012) and hyperarid soils (Atacama Desert; Orlando et al. 2012), and BSCs (Abed et al. 2013). Studies suggest that semiarid soils have a higher denitrification potential (based on the presence of nirK and nirS genes) than hyperarid soils attributable to its higher soil moisture and nutrient content (Orlando et al. 2012). Denitrification rates in BSCs are likely to surpass those in uncrusted soils due to the formation of SOC in crusts (Barger et al. 2016; Brankatschk et al. 2013), however, contradictory results exist to support this hypothesis (see Barger et al. 2016 for full review on denitrification in BSCs).

7.3 Phosphorus in Dryland Soils

Phosphorus (P) is one of the less-abundant macronutrients in the lithosphere (0.1% of total) and then considered as a limiting nutrient (Bate et al. 2008), thus turnover of organic phosphorus (P_o) is critical for community structure, species diversity, and ecosystem processes (Runge-Metzger 1995; Sardans et al. 2012). P concentration is fundamental in the cycling of nutrients such as soil organic matter (SOM), nitrogen (N), and C (Stevenson and Cole 1999; Mackenzie et al. 2002; Vitousek et al. 2010; Griffiths et al. 2012; George et al. 2018), and vital to cellular organization (phospholipids) and genetic material (nucleic acids), intracellular signaling molecules, and for metabolism and energy transfer (ATP; Ruttenberg 2003; Butusov and Jernelöv 2013).

In mesic terrestrial environments, soil P is derived from three sources: (1) chemical and biological weathering of bedrock; (2) aeolian dust deposition; and (3) decomposition of biomass (Walker and Syers 1976; Lajtha and Schlesinger 1988; Okin et al. 2004; Turner et al. 2007; Yang and Post 2011; Porder and Ramachandran 2013). In most soils, P is the result of geochemical weathering of parent material (Lajtha and Schlesinger 1988), with relatively small inputs from aeolian deposition (Okin et al. 2004; Selmants and Hart 2010). Plants and microbes incorporate P into biomass and return it to the soil in organic forms, which can then be recycled by phosphatase enzymes to release inorganic phosphate for biological uptake (Walker and Syers 1976; Cross and Schlesinger 1995; Turner et al. 2007). Contrariwise, the input of P into dryland soils is primarily the result of atmospheric dust deposition (see Belnap 2011 for a comprehensive description), followed by biological weathering of parent materials (Reynolds et al. 2006; Belnap 2011), and then topsoil (0–15 cm) typically contain high concentrations of total P (Jobbagy and Jackson 2001; He et al. 2014).

7.3.1 P Stocks and Redistribution by Biological Processes in Drylands

Phosphorus stocks in dryland soils range from ~200 to 1200 mg P kg^–1 (Tiessen et al. 1984; Turner et al. 2003a; Plaza et al. 2018a). More than 50% of the total P consists of labile inorganic and apatite P, that are mostly unavailable for plant or microbial uptake (Cross and Schlesinger 2001; Jones and Oburger 2011). Since P_o is mostly absent or at low concentrations, variation in P fractions can be closely related to the P_i content of parent materials (Neff et al. 2006; Buckingham et al. 2010). In addition, P_i is more likely to be found in mineral-associated forms such as oxides (e.g., aluminum and iron), clays, or carbonates (Sinsabaugh et al. 2008). In fact, calcium (Ca) minerals (e.g., calcium carbonate-and phosphate) are the predominant reservoir of P (Lajtha and Schlesinger 1988; Cross and Schlesinger 1995; Sims and Pierzynski 2005; Belnap 2011) in drylands due to reduced water inputs and high evapotranspiration.

The ability of P to change from one fraction to another is mainly controlled by mineralogy (Cross and Schlesinger 2001; Lajtha and Schlesinger 1988; Neff et al. 2006). Several different minerals play important roles in the stabilization and release of P (Buckingham et al. 2010) in the soil matrix. For example, studies have shown that carbonates sorbs P moderately due to its positive electrostatic charge, although the association can be reversed by changes in pH (Tiessen and Moir 1993), where goethite quickly sorbs P via ligand exchange until the sorption requirements of the mineral are met (Goldberg and Sposito 1985). Some minerals, such as aluminum (Al) and iron (Fe) oxides, may stabilize P through several mechanisms (Buckingham et al. 2010). These oxides can sorb P onto mineral surfaces that are reversible or may exchange ligands that sequester P into the physical structure of the oxides (Sollins et al. 1988). Amorphous minerals tend to sorb P more rapidly than crystalline materials (Ryan et al. 1985) and form tighter bonds with P because of their larger surface area (Carreira et al. 2006). Ultimately, the type and intensity of P-mineral bonds are important characteristics that determine its mobility and plant availability in soils (Buckingham et al. 2010).

In un-weathered or moderately weathered soils with neutral to alkaline pH, calcium-phosphates are the main supply of P_i. In contrast, Al and Fe phosphates and P_i bound and/or occluded by Al and Fe oxy(hydr)oxides predominate in acidic and more progressively weathered soils (Sims and Pierzynski 2005). In neutral and acidic soils, Al and Fe oxides and hydroxides exert a great impact on P availability given that various identified Fe and Al phosphates (e.g., wavellite, variscite, and strengite) are generally rare in occurrence (Harris 2002). In addition, the increased positive surface charge of the oxides forms strong covalent bonds (chemisorption) with the negatively charged P under acidic conditions, rendering it rather recalcitrant to exchange reactions (Jones and Oburger 2011). Nevertheless, changes in pH might directly or indirectly affect the oxides’ surface potential and consequently P_i solubility. Also, LMW organic anions (e.g., gluconate, oxalate, etc.) released by P solubilizing organisms are capable of competing with P_i for sorption sites (Jones and Oburger 2011).

P_o influences P availability by contributing to the labile P_i pool, which is important to NPP as mentioned previously (Cross and Schlesinger 2001). Certain fractions of P_o are relatively labile and accessible for biological uptake (Haas et al. 1961; Wild and Oke 1966; Dalal 1977), while unavailable P_o is mineralized to labile P_i by simple autolysis or enzymatic phosphorylation (Cosgrove 1977). On average, P_o can comprise 30–65% of total P in mineral soils, while P_o approaches up to 90% in organic soils (Dalal 1977; Jones and Oburger 2011). The main P_o compounds in soil include inositol phosphates (>80%), phospholipids (0.5–7.0%), and nucleic acids (<3%; Dalal 1977; Quiquampoix and Mousain 2005; Turner et al. 2002; Jones and Oburger 2011). Other, less abundant forms of P_o include sugar-P (Anderson 1961), monophosphorylated carboxylic acids (Anderson and Malcolm 1974), and teichoic acids (Zhang et al. 1998).

The redistribution of P into various mineral and chemical fractions is impacted by mechanical and biological weathering, biological uptake and cycling, adsorption onto secondary minerals or soil constituents, or leaching (Walker and Syers 1976; Tiessen et al. 1984; Ruttenberg 2003). These P pools are typically represented as soluble phosphate, P_o, and primary and secondary mineral fractions that vary in their potential mobility, relative quantities, and availability to biota (Hedley et al. 1982; Tiessen et al. 1984; Kuo 1996). The horizontal (e.g., wind, water, wildlife, livestock, human activities) and vertical (e.g., animals, insects, and plants) redistribution of soil P in dryland systems is imperative for plant productivity. Redistribution of P can determine the community, biomass, and distribution of a given plant community, which in turn affects animal distributions (Belnap 2011). For a comprehensive review on redistribution of soil P in drylands see Belnap (2011).

7.3.2 Abiotic and Biotic Control of P Cycling

Biogeochemical cycling of P in drylands is ultimately controlled by precipitation and temperature (Belnap 2011; Hou et al. 2018). The timing, intensity, and amount of water input directly affect the availability of P by influencing rates of geochemical reactions, ion diffusion, and biological activity (Belnap 2011). Because precipitation in arid environments is highly variable and inherently low, pulses of P-releasing activities vary on both spatial and temporal scales (Belnap 2011). While precipitation affects soil P leaching, temperature can directly enhance soil P sorption and desorption (Barrow 1983). Climate may also affect soil P availability through additional factors including: (1) different soil P forms; (2) soil properties such as soil particle size, pH, and SOM; and (3) plant and soil microbial activities (e.g., P uptake and return; Hou et al. 2018). For instance, high temperature facilitates the transformation of labile P and secondary mineral P to occluded P in soil (Barrow 1983; Siebers et al. 2017), whereas high precipitation may drive the loss of secondary mineral P or even occluded P in very humid environments (annual precipitation >2000 mm; Austin and Vitousek 1998). The interaction of precipitation and temperature can also affect the release of bound P (see Belnap 2011). Essentially, cool and wet conditions result in greater production of carbonic acid (H₂CO₃) that causes the soil pH to decrease, thus dissolving carbonates and increasing the transition of solid-phase P to solution-phase P (Lajtha and Schlesinger 1988; Magid and Nielsen 1992; Jungk and Claassen 1997; Miller et al. 2006a, b).

While the availability of P_i in dryland soils is mainly regulated by the dissolution properties of P-bearing minerals, the availability of P derived from P_o is largely governed by microbial activity (Jones and Oburger 2011). Soil microorganisms can solubilize P via three mechanisms including: (1) the active or passive release of protons, CO_2, and secondary organic metabolites (e.g., sugars, organic acid anions, amino acids, siderophores, enzymes, phenols); (2) the release of extracellular phosphatase enzymes (biochemical P_o mineralization); and (3) the release of P_o during substrate degradation (biological P_o mineralization) (Mcgill and Cole 1981; Jones and Oburger 2011). P incorporated into microbial biomass, to allow the use of organic C and root exudates for energy (Wu et al. 2007), may be temporarily immobilized but remains in a bioavailable form that can be released via microbial turnover (re-mineralization; Jones and Oburger 2011).

Approximately 1–50% of soil bacteria and 0.5–0.1% of soil fungi can be classified as P-solubilizing microorganisms (PSM; Kucey et al. 1989; Gyaneshwar et al. 2002); however, their population and activity vary with environment and depend upon various environmental factors (Chatli et al. 2008). P-solubilizing organisms are capable of hydrolyzing P_o and P_i compounds from insoluble P (Kalayu 2019). Among the PSM group, species from bacterial genera (Bacillus, Pseudomonas, Rhizobium, Burkholderia, Achromobacter, Agrobacterium, Microccocus, Aereobacter, Flavobacterium, and Erwinia), fungal general (Penicillium, Aspergillus, and Trichoderma), and actinomycetes (Streptomyces and Streptoverticillium) are significant in solubilizing phosphate (Rodriguez and Fraga 1999; Whitelaw 2000; Kumar et al. 2018; Kalayu 2019). Although the number of bacteria in soil classified as PSM generally outnumber those of fungi, the fungal isolates generally exhibit a greater P-solubilizing capacity (Banik and Dey 1982; Gyaneshwar et al. 2002). The mechanisms and processes involved in P mobilization by PSM in soil are discussed elsewhere (Jones and Oburger 2011; Richardson and Simpson 2011).

Microbial immobilization of P constitutes a significant component of the total soil P and is generally equivalent to, or exceeds that held in plant biomass (Richardson and Simpson 2011). Concentrations of microbial P in dryland soils typically account for 2–4% of total soil P (Lajtha and Schlesinger 1988; Xu et al. 2013; Perroni et al. 2014). Importantly, microbial P is a highly dynamic pool of soil P and release of P (i.e., orthophosphate and in organic forms) during the microbial biomass turnover is subject to significant change in response to environmental factors such as seasonal conditions, soil temperature and moisture, and C availability (Patra et al. 1990; He et al. 1997; Butterly et al. 2009; Richardson and Simpson 2011).

Carbon dynamics in soil have been closely linked to microbial biomass P (Achat et al. 2009) where immobilization of P is highly regulated by C availability (C:P ratio; Spohn and Kuzyakov 2013; Heuck et al. 2015). Microbial phosphatases and other enzymes are involved in organic C mineralization to promote the availability of P and C from organophosphorylated compounds (Spohn and Kuzyakov 2013; Heuck et al. 2015). Thus, P immobilization and transformation is inherently coupled, to a certain extent, with C mineralization (Luo et al. 2020). As soil microbial activity increases with the availability of C (i.e., SOM), immobilized P is released to increase the available P during microbial biomass turnover (Yang et al. 2010; Haripal and Sahoo 2014). When C sources become limited microbial biomass P will subsequently decrease (low C:P ratio; Yang et al. 2010; Haripal and Sahoo 2014). Given that the turnover of microbial P is largely driven by the C availability it is of particular importance in the rhizosphere (Richardson and Simpson 2011). As orthophosphate diffuses through the rhizosphere/mycorrhizosphere (Jakobsen et al. 2005) it is in direct competition for uptake and immobilization by microorganisms. Subsequently, the rate of release of P from microorganisms or microbial biomass turnover time within the rhizosphere will have major implications for P availability to plants (Richardson and Simpson 2011).

7.4 Role of Biological Soil Crust in Nutrient (C, N, and P) Cycles

Biological soil crusts (BSCs) or biocrust consist of photoautotrophic (green microalgae and cyanobacteria) and heterotrophic (bacteria and fungi) organisms, stable layers predominantly on bare soil surfaces in which inorganic particles are bound together by sticky extracellular polymeric substances (EPS; Belnap et al. 2001). The importance of biocrusts in the functioning of ecosystems in arid environments is well documented (Veste et al. 2001; Belnap et al. 2003; Maestre et al. 2011; Chap. 3).

It is estimated that BSCs are responsible for global C fixation of ~7% of the terrestrial vegetation, and for N fixation of ~50% of terrestrial biological N fixation (Elbert et al. 2012). Much of the C that is photosynthetically assimilated by biocrusts is released to the underlying soil shortly after fixation, thereby increasing the total amount of C and organic matter in soil (Pointing and Belnap 2012). Biocrusts also regulate the temporal dynamics of soil CO₂ efflux and net CO₂ uptake (Castillo-Monroy et al. 2011; Wilske et al. 2008, 2009). Studies have shown that temporal increase in soil moisture (i.e., higher and more intense precipitation frequencies) significantly increases C inputs in BSCs (Wilske et al. 2008; Huang et al. 2014), although the moisture levels at which organisms become photosynthetically active are highly variable (Belnap et al. 2004). Concurrently, higher moisture levels and temperature induce CO₂ emissions (Maestre et al. 2013) by several factors:

• (i) Infiltration of water may physically displace the CO₂ accumulated in soil pore spaces during the dry season.

• (ii) Rewetting triggers microbial activity in both biocrusts and underlying soil matrix that results in increased respiration rates.

• (iii) Water addition may induce CO₂ release from microbial biomass pool (Li et al. 2018).

Biocrusts may also influence the activity of C-related soil enzymes, depending on the environmental conditions (e.g., β-glucosidase; Bowker et al. 2011; Yuan and Yue 2012; Miralles et al. 2013). When moisture and organic C content increases enough, sufficient resources are available to support higher microbial biomass and thus higher enzyme activity. High organic C content also allows existing enzymes to be stabilized by their absorption to SOM (Yuan and Yue 2012). Conversely, low moisture content and organic C result in relatively low microbial growth, enzyme stabilization, and activity (Yuan and Yue 2012). BSCs type plays an important role in C fixation at sites (Sancho et al. 2016; Maier et al. 2018). For example, Miralles et al. (2018) have demonstrated that late-successional biocrusts (i.e., lichens and mosses) had higher gross photosynthesis than early-successional biocrusts (cyanobacteria) in two semiarid ecosystems. Similar results were obtained by Housman et al. (2006) for the Colorado Plateau and Chihuahuan Desert where late-successional biocrusts (i.e., cyanobacteria [Nostoc and Scytonema] and lichens [Placidium and Collema]) had higher C fixation rates (1.2–1.3-fold) than early-successional biocrusts (cyanobacteria [Microcoleus]). Estimates for annual C inputs range from 0.4 to 2.3 g C m^–2 year^–1 for early successional crusts to 12–37 g C m^–2 year^–1 for late-successional crusts (Evans and Lange 2003; Li et al. 2012b; Yan-Gui et al. 2013), representing 1% of the NPP of terrestrial vegetation (~56 Pg year^–1; Zhao et al. 2005). In terms of dryland systems, biocrusts account for ~9% of the total NPP, corresponding to ~0.07 Pg year^–1 compared to the total NPP of ~0.8 Pg year^–1 (Zhao et al. 2005; Elbert et al. 2012; Sancho et al. 2016). In terms of C release, studies have found that 40–60% of soil respiration in dryland ecosystems were attributable to (algal) biocrust-dominated sites (Castillo-Monroy et al. 2011; Zhang et al. 2013). However, soil respiration is highly dependent on the type of crust, temperature, and precipitation (Tucker and Reed 2016; Zhang et al. 2016b; Guan et al. 2018; Li et al. 2018).

Given that most BSC components can fix C, the availability of P increases as organic matter accumulates in soil (Belnap 2011). The availability of P is further intensified through the secretion of extracellular phosphatases, organic acids (e.g., oxalic acid, citric acid, and malic acid), and metal chelators (e.g., siderochromes; Lange 1974; Belnap 2001a, 2011; Crain et al. 2018). Most BSC organisms (soil and hypolithic cyanobacteria, green algae, lichens, and mosses) contain phosphatases in their cell walls and mucilaginous sheaths, but they also release extracellular phosphatases into the surrounding soil (Belnap 2011). Phosphatases hydrolyze organic phosphates to subsequently release P (Turner et al. 2003b; Nannipieri et al. 2011; Baumann et al. 2017; Crain et al. 2018) which can then be immobilized by microbes, transferred to plant host roots, or stabilized by humic substances (Sinsabaugh 1994; Lindahl et al. 2005). As phosphatase activity is highly correlated with SOM, which occurs at low concentrations in dryland soils, reduced phosphatase activity is expected compared to more mesic environments (Sinsabaugh et al. 2008).

In addition to nutrient cycling, BSCs control the water balance (Warren 2003; Chamizo et al. 2016) by regulating the following factors (Chap. 3):

• (i) Water infiltration and runoff (Belnap 2006; Chamizo et al. 2012, 2016; Zaady et al. 2013)

• (ii) Increase soil moisture retention

• (iii) Protect soil surface aggregates

• (iv) Prevent soil erosion by water or wind

• (v) Facilitate colonization by vascular plants (Belnap 2003)

• (vi) Contribute to the stabilization of sand dunes (Eldridge and Greene 1994; Belnap 2002)

The presence of BSCs can also alter soil surface conditions in ways that affect the suitability of the habitat for other organisms (Bowker et al. 2005, 2006). However, their metabolic activity is strongly linked to moisture availability (Sancho et al. 2016). Due to low and highly variable precipitation pulses in drylands, biocrusts are sporadically active (Johnson et al. 2012; Yu et al. 2014; Sancho et al. 2016; Fernandes et al. 2018; Miralles et al. 2018). Nevertheless, their distinguishing characteristics highlight their overwhelming ecological importance in dryland environments, particularly in element cycling where the C, N, and P cycles are interlinked (Delgado-Baquerizo et al. 2013).

7.5 Influence of Hydration-Desiccation Pulses on Nutrient (C, N, and P) Cycles

Infrequent precipitation events occur in drylands; thus, soil surfaces are most often dry. During desiccation periods, nutrients accumulate on the soil surface due to dust deposition and the degradation and/or death of organisms from UV exposure (Belnap 2011). When precipitation occurs at higher temperatures, microbial diversity and function tend to rapidly increase thus affecting the rate of nutrient cycling in soils (Belnap 2011; Bell et al. 2014; Montiel-González et al. 2017; Chap. 11). Should precipitation occur at the proper time to stimulate annual plant activity for a few weeks to months, subsequent C inputs will stimulate soil microbial activity, including processes that release bound P (Whitford 1999). However, when a small amount of precipitation occurs (<5 mm rain) there is a shallow penetration of water into the soil, and substantial buildup of nutrients at the surface, that elicit more responses from soil surface organisms and their associated processes than vascular plants. The temporal dynamics of water use between microbial and plant communities can result in decoupling between their responses to precipitation and nutrients (Stursova et al. 2006; Belnap et al. 2004), but also in the accumulation of nutrients (hydrolyzable P_o and C) at the soil surface between large precipitation events (Whitford 2002; White et al. 2004).

In that sense, much attention has been given to hydration-desiccation cycles in drylands and the impact of rewetting on microbial communities and nutrient cycling (emphasis on C) in both laboratory conditions and field studies (Austin et al. 2004, Bell et al. 2014; Frossard et al. 2015; Montiel-González et al. 2017; Schimel 2018; Chap. 11). As soils dry, diffusion rates decline and semi-soluble materials precipitate (Schimel 2018) causing water stress. Microbes will increase their intracellular solute concentration to compensate for the extracellular concentration and counterbalance the increased osmotic pressure (Stark and Firestone 1995; Bell et al. 2008). This high concentration of solutes results in an inhibition of the enzymatic activity and therefore decreases cellular activity (Batlle-Aguilar et al. 2011). As a consequence, microbial activity and respiration is reduced (Stark and Firestone 1995; Austin et al. 2004; Schimel 2018), and a portion of microbial biomass is killed under such conditions (Austin et al. 2004). However, the magnitude of the respiration decrease may vary; in some cases, substantial reductions in soil moisture have limited effects on respiration (Lu et al. 2017).

Rewetting of soils after a drought period can result in large pulses of CO₂, also known as the “Birch effect” (Birch 1958). The CO₂ pulse can be many times greater than the basal respiration level (Schimel 2018), although the size of the pulse may largely depend on the labile C soil pools and the soil’s history of physical disturbance (Austin et al. 2004; Schimel 2018). The exact mechanism that drives the Birch effect remains unclear and contrasting patterns emerge from rewetting experiments (see Schimel 2018 for a comprehensive overview). A study by Fierer et al. (2003) suggests that during each rewetting cycle, carbon is immediately lost from the (dead) microbial biomass and labile SOM is simultaneously mobilized from the soil matrix. The mobilized material may then be utilized by microbes to replace lost biomass C and to promote additional growth (Schimel 2018). Therefore, through multiple cycles, the C mobilized may ultimately be from stable soil C, but within each cycle, the C originates immediately from microbial biomass (Schimel 2018).

In addition to C, wet–dry cycles have also important consequences on N cycling processes. In drylands covered with BSCs, water pulses have a temperature-dependent impact: cold desert soils remain moist for longer periods of time than hot desert soils, thereby increasing the N inputs into soils (Austin et al. 2004). Since rates and duration of nitrogenase activity are mainly controlled by water availability (Hartley and Schlesinger 2000; Belnap 2002; Hartley and Schlesinger 2002; Wu et al. 2009; Zhou et al. 2016), N fixation in drylands may start within a few hours following a wetting event (Abed et al. 2010) as organic matter increases due to primary production (Wierenga et al. 1987). For example, nitrogenase activity was detected within 2 h after wetting dry cyanobacteria-dominated crusts (i.e., Nostoc, Scytonema, and Microcoleus) from Sahel, Niger (Malam Issa et al. 2001). In the Great Basin, nitrogenase activity was observed within 3 days after wetting cyanolichen crusts (Jeffries et al. 1992).

Wetting of dry soils also induces NO emissions, which can reach 400-fold compared to dry soils (Austin et al. 2004). Likewise, water pulses have a critical effect on denitrification where rates can increase significantly after precipitation events (Austin et al. 2004 and references thereof). Even smaller water pulses, either by rainfall, fog, or dew, following a large precipitation event can readily activate soil and BSC microorganisms to respire, photosynthesize, and perform other metabolic processes (Belnap and Lange 2001; Belnap 2003; Veste et al. 2008). N formed by metabolically active BSCs, mainly as NH₄⁺, may leak to the environment increasing nutrient availability for nearby vegetation and microorganisms (reviewed in Belnap 2001b). Increased N will also lead to higher N loss through nitrification and denitrification processes as discussed above (Zaady 2005; Barger et al. 2005; Johnson et al. 2007; Strauss et al. 2012; Brankatschk et al. 2013; Liu et al. 2016).

An important aspect to consider in soil respiration is that of microbial community composition. Bacteria and fungi are ecologically and physiologically distinct groups, support different soil food-chains (De Vries et al. 2006) and differ in their C use efficiency (Sakamoto and Oba 1994). Studies have demonstrated that fungi have a higher resistance to drought and that fungal growth is largely unresponsive to drying-rewetting cycles (Bapiri et al. 2010; Evans and Wallenstein 2012; Manzoni et al. 2012; Barnard et al. 2013; Canarini et al. 2017). However, new evidence suggests that fungal communities are more strongly influenced by drying-wetting cycles than previously considered and that communities shift toward a more Ascomycota-dominated population (Hicks et al. 2019; Liu et al. 2019). In contrast, bacterial communities are more responsive to wetting pulses and may show an immediate and linear growth response following rewetting, or bacterial growth starts exponentially after a prolonged lag-period of no growth (Meisner et al. 2013, 2015). This causes changes in the fungal and bacterial biomass, as well as the fungal:bacterial (F:B) ratio, which ultimately influences the C use efficiency and C loss from soils (Fig. 7.1; Canarini et al. 2017).

Fig. 7.1

Conceptual model of C dynamics during a drying-rewetting cycle summarizing the observed patterns obtained in a meta-analysis by Canarini et al. (2017). Fungi maintain higher respiration rates during the drying period, while bacteria show higher resilience and faster growth rates to return to pre-drought conditions when diffusion is restored (rewetting), causing a burst in respiration. Carbon-rich soils may boost this mechanism due to a higher fungal population, and by a greater release of available substrates (accumulated during the drying phase) to stimulate the bacterial community. A drought intensity threshold must be reached to initiate this response, possibly by regulating diffusion and therefore modulating the connection between available substrate and microorganisms. Adapted from Canarini et al. (2017)

7.6 Impact of Climate Change on Nutrient Cycling

Climate models estimate that global warming will likely aggravate dryland expansion due to increases in evaporative demand and a global hydrological cycle with longer and more severe droughts (Schlaepfer et al. 2017; Koutroulis 2019). Of particular concern is the decrease in soil moisture that can result in: (1) changes in vegetation due to shifts in plant functional types (Harrison et al. 2015), woody plant mortality (Allen et al. 2010), and encroachment (D’Odorico et al. 2012), and resistance of some vegetation types (Craine et al. 2013); and (2) rise in temperatures within 3.0–4.0 °C (IPCC 2014; Huang et al. 2017; Koutroulis 2019).

Also, increase in aridity along chrono- and climosequences has been shown to cause a decline in the concentration of both the total and the most biologically available forms of C and N, but an increase in biologically available forms of P, hence distorting soil C, N, and P cycles (Finzi et al. 2011; Delgado-Baquerizo et al. 2013; Feng et al. 2016). This results in an overall decrease in C:P and N:P ratios and decoupling of C and N cycles from P (Fig. 7.2; Evans and Burke 2013; Delgado-Baquerizo et al. 2013; Wardle 2013). Such changes to the environment and uncoupling of nutrient cycles may have detrimental effects on ecosystem function (Delgado-Baquerizo et al. 2013). Reduced availability of C and N may skew C:N:P stoichiometry in soil, constraining plant and microbial activity and diversity (Finzi et al. 2011; Peñuelas et al. 2012). This may negatively impact biogeochemical reactions that control key ecosystem functions (e.g., primary production, respiration, and decomposition) and services (e.g., food production and C sequestration; (Dodds et al. 2004; Schimel and Bennett 2004; Schimel 2010; Finzi et al. 2011; Delgado-Baquerizo et al. 2013; Maestre et al. 2016; Lal 2019), that can jeopardize human well-being and nature conservation (Schlaepfer et al. 2017).

Fig. 7.2

Increase in aridity causes element cycles (C, N, and P) to decouple. Aridity is predicted to increase in many dryland ecosystems worldwide due to climate change. Delgado-Baquerizo et al. (2013) reported that, as aridity increases, available soil carbon (C) and nitrogen (N) decline, whereas available soil phosphorus (P) increases. This is a result of the impairment of biological processes that contribute to the C and N levels, and of an increase in the relative importance of abiotic processes that contribute to P availability. Adapted from Wardle (2013)

Changes in temperature, soil moisture, and vegetation structure and composition will likely modify the C sequestration capacity in high latitude drylands via direct warming effects on photosynthesis and decomposition of primary producers (Sjögersten and Wookey 2009), affecting the amount of biomass SOC stored in soils. In addition, evidence suggest that increase in temperature will affect biocrust communities that will consequently reduce their capacity to act as a carbon sink (Maestre et al. 2013; Darrouzet-Nardi et al. 2015; Escolar et al. 2015). For example, warming will reduce biocrust cover, richness, and diversity (Maestre et al. 2013; Ladrón de Guevara et al. 2018; Lafuente et al. 2018; Eldridge and Delgado-Baquerizo 2019), and communities will shift from late (i.e., dominated by mosses and lichens) toward early successional states dominated by cyanobacteria (Ferrenberg et al. 2015). This has critical implications for C sequestration and ecosystem functioning as a reduction in biocrust cover and the formation of early successional biocrusts will lessen the C storage capacity (Housman et al. 2006; Ferrenberg et al. 2015; Darrouzet-Nardi et al. 2018) of soils. This decrease may act synergistically with other warming-induced effects, such as the increase in CO₂ efflux (Maestre et al. 2013; Darrouzet-Nardi et al. 2015, 2018; Escolar et al. 2015) and changes in microbial communities, to alter C cycling in drylands and ultimately reduce soil C stocks in the mid to long term (Maestre et al. 2013).

Importantly, drylands are sensitive to land degradation and desertification which dramatically reduce the amount of SOC stored in soils (Lal 2004a; Serrano-Ortiz et al. 2012). Previous studies have estimated that the historic SOC loss from drylands due to desertification ranges between 13 and 29 Pg C (Ojima et al. 1995; Lal et al. 1999; Lal 2001). Additionally, changes in vegetation structure induced by overgrazing promote C losses that may significantly change both plant and SOC pools (Gaitán et al. 2017; Abdalla et al. 2018). As SOC has a major impact on soil physical structure and ecosystem function (e.g., nutrient retention, water storage, pollutant attenuation), its reduction can lead to reduced soil fertility and further land degradation (Abdalla et al. 2018). The magnitude of this loss underscores the importance of appropriate management systems and best strategies to restore, preserve, or increase dryland SOC to enhance ecosystem functions and services, and mitigate climate changes (Plaza et al. 2018b).

Climate change will also significantly alter N cycling processes in drylands as NO_x emissions are related to temperature (McCalley and Sparks 2009; Suddick et al. 2013). Optimal temperatures for N processes (e.g., N fixation) generally range between 20 and 30 °C (Trolldenier 1982; Montañez et al. 1995). With global temperatures rising, BNF and thus N input will consequently induce larger NO_x emissions in hot deserts where summer temperatures already exceed 40 °C (Makhalanyane et al. 2015). In addition, higher N losses will further contribute to soil infertility, therefore unable to support most plant life. Although some climate models predict higher summer precipitation events for desert areas (e.g., arid regions of central Asia), the combination of moisture and heat would greatly enhance N losses. One should also consider that most climate models only consider biological factors to predict N gases losses from soils and rarely account for abiotic impacts on the N budget (McCalley and Sparks 2009).

P cycling will be also strongly influenced by climate change, as P availability is directly and indirectly impacted by both temperature and precipitation in drylands (Belnap 2011). Plants whose distribution is limited by available P will likely undergo spatial shits in habitat (Belnap 2011). Consequently, biological cycling of P will be reduced due to manifold factors: (1) a decline in soil moisture will not only negatively affect abiotic processes by slowing the release of bio-unavailable P, but also biotic processes (Belnap 2011). Reduced soil moisture will decrease microbial abundance and activity (Sardans et al. 2006), resulting in reduced acidification and production/excretion of enzymes, chelators, and other compounds to release bound P (Belnap 2011). Lower plant biomass, including root biomass, will further reduce soil acidification and root exudates to release bound P (Delucia et al. 1997; Li and Sarah 2003); (2) low soil moisture will also affect phosphatase activity as their effectiveness is more dependent on water availability than substrate availability (Sardans et al. 2008); (3) infrequent hydration-desiccation events will cause less P to be released from physical processes or from dead microbes (Belnap 2011); (4) severe droughts and higher temperatures will increase plant resorption of P, resulting in less P in detritus material (Killingbeck and Whitford 2001; Sardans et al. 2006; Luo et al. 2018); and (5) reduced plant cover can intensify dust emissions (Field et al. 2010; Duniway et al. 2019), resulting in substantial loss of many plant-essential nutrients including P, Na, K, and Mg from soil (Li et al. 2007; Belnap 2011; Katra et al. 2016).

7.7 Conclusion

Nutrient cycling in dryland soils and BSCs is a complex, yet active, process largely controlled by water availability (i.e., the size and frequency of precipitation events). Other abiotic factors such as temperature, C content and light, but also microbial species composition determine the rate by which nutrients are cycled. As aridity in drylands increases with climate change, the coupling between nutrient cycles will weaken and many arid environments can observe abrupt declines in organic C and total N content, while inorganic P will accumulate. Such changes are likely to have adverse effects on key ecosystem functions (e.g., primary production, respiration, and decomposition) and food services (e.g., food production and C storage). As such, comprehending the role and responses of microbial communities in drylands (soil and BSCs) can offer vital insights/predictions into future C and N fluxes, with important implications for future sustainable management and conservation policies.