The Essaouira region in Morocco seems to be a suitable study area due to the importance of the living dune ecosystem (latitude of 31° 30′ 0.00″ N; longitude of −9° −45′ −36, 00″ W) and the heavy invasion process conducted by A. cyanophylla. Three sites were randomly chosen to find out how A. cyanophylla changes soil chemical and biological fertility. They were named site 1, site 2 and site 3. On 4 March 2019, five A. cyanophylla plants were randomly picked at each site. At a depth of 10–20 cm, soils directly influenced by the roots (rhizosphere) were collected. Additionally, the bare soil (served as a control) was gathered well away from any plant roots at each location. Fifteen rhizosphere soil samples and three bare soil samples were collected. A composite soil was produced for each of the three sites by mixing the five samples homogeneously. The bare soils were also combined to generate a homogeneous sample (control).

The physicochemical properties of the examined soils are given in Table 1. Initially, all soil samples were air-dried and sieved at 2 mm. Afterward, an aliquot of soil was suspended in distilled water (1:2.5, w/v) before being stirred for 30 min using a mechanical agitator. The pH of each soil was measured by a calibrated digital pH meter (JP Selecta pH-2006). The electric conductivity of the soil was measured after magnetic stirring for 20 min of 10 g of soil in 50 ml of distilled water; the measurement was carried out using a conductivity meter calibrated with a KCl solution (0.001N).

The organic matter was determined according to Anne’s method as described by Genin et al. (2017), which is based on the oxidation of soil organic carbon by potassium dichromate in an acidic medium and the back titration of the excess potassium dichromate by a solution of Mohr’s salt in the presence of diphenylamine sulfonate as an indicator. After titration, the percentage of total organic carbon (%TOC) was calculated according to the following formula: % TOC   =   3.9   V c − V s m ∗ V c (1) where: Vc (cm³) = Control volume, Vs (cm³) = sample volume and m (g) = soil dry mass.

The soil organic matter content (%OM) was calculated as follows: % OM   = %TOC ∗  100 58 (2)

To assess the mineral content of A. cyanophylla soil, 2 g of rhizospheric and bare soil were digested in sulfuric acid solution at 600°C for 6 h. Ash was mixed with 2 ml of HCl (10N), agitated, evaporated on a hot plate, and collected in 100 ml of distilled water. The contents of K⁺, Na⁺ and Ca²⁺were determined according to the method of Brown and Lilleland, (1946) using a flame spectrophotometer. The total nitrogen was determined using the Kjeldahl method (Bremner, 1960). The available phosphorus was determined using the molybdenum blue colorimetric technique (Wei et al., 2009). The optical densities of the solutions were measured at 880 nm using an S-22 UV/Vis spectrophotometer (Boeco, Germany).

Isolation of arbuscular mycorrhizal fungi spores from rhizosphere soil was performed according to the method described by (Brundrett et al., 1996), based on the extraction of spores by wet sieving and decanting of soil followed by centrifugation in a sucrose solution. So, 100 g of rhizosphere soil was suspended in 1 L of tap water. The mixture was then stirred and then left to settle. After decantation, the supernatant was poured through a series of sieves ranging in size from 800 μm to 50 μm to remove the largest particles of organic matter from the soil while still retaining spores of different sizes. The operation was repeated several times to recover as many spores as possible. Spores suspension retained in sieves was suspended in distilled water and centrifuged for 5 min at 2000 rpm, the supernatant and floating debris were removed to add 20 ml of sucrose solution (65%) to the pellet for 1 min centrifugation at 2000 rpm to separate the spores of denser soil components. The supernatant was filtered through a filter paper under a vacuum, and the spores were collected on Petri dishes for enumeration using a binocular magnifying glass (Magnification:×40).

Spore extraction was based on the wet sieving and decantation of the soil samples (Gerdemann and Nicolson, 1963) and the concentration of the spores by centrifugation on a sucrose solution (Brundrett et al., 1996). The isolated spores were sorted under the binocular (Magnification ×40) according to morphological characters (color, size, shape). Other morphological parameters were used under a microscope after mounting the spores between slide and coverslip, such as the diameter of the spores, the lamellar structure of the spore wall, and the form of attachment of the hyphae to the spore. Finally, recovered spores were stored in a polyvinyl-lacto-glycerine (PVLG) medium until use (Estaún et al., 1997). Spores with similar morphological traits were grouped in a homogeneous group called a morphotype. The relative abundance of each morphotype was calculated as the number of spores of that morphotype divided by the total number of spores from each soil (El kinany et al., 2018).

Thinning and coloring of roots were carried out following the modified method of Phillips and Hayman (1970). The roots were washed under a gentle stream of water to remove attached soil, organic matter, and foreign roots. A 10% KOH solution was used to bleach the roots for 60 min at a temperature of 90°C. The samples were then rinsed and acidified by adding a few drops of 5% lactic acid, which neutralized the remaining KOH. The thinned roots were stained with a 0.05% acidic trypan blue solution, diluted in lactoglycerol (1/3 water, 1/3 glycerol and 1/3 lactic acid) for 15 min at 90°C. The colored roots were then mounted between the slide and coverslip and observed using a Microscope. The mycorrhizal frequency represents the number of mycorrhizal roots in the total examined roots. While the mycorrhizal colonization provides information on the total volume of the root colonized by different fungal structures. It was estimated according to five classes of colonization 0 (non mycorrhized at all), 1 (trace of mycorrhization), 2 (less than 10% colonization), 3 (between 11 and 50% colonization), 4 (between 51 and 90%), 5 (more than 91% colonization) (McGonigle et al., 1990; Brundrett et al., 1996). The volume occupied by each different mycorrhizal structure (Mycelium, vesicles, arbuscules) was also estimated by microscopic observations. M% = ( I n 5   +  II n 4   +  III n 3   +  IV n 2   +  V n 1   +  VI n 0 ) (3) where n₅ is the number of fragments with index 5, n₁ is the number of fragments with index 1, and I is the frequency assigned to the index 5.

The absolute mycorrhization intensity (Mycorrhization intensity among mycorrhized fragments) is calculated as follows: m% = M% Mf   ∗   T f (4) where Tf = Total fragments and Mf = Mycorrhized fragments.

The Most Probable Number (MPN) method was performed to determine the mycorrhizal infectious potential of a soil which exhibits the ability of that soil to initiate the formation of mycorrhizal associations in a test plant from a quantity of inoculum present in rhizosphere soil in the form of spores, mycelium and root debris bearing vesicles (propagules) (Ouahmane et al., 2012). For this, the five samples of rhizospheric soils taken were mixed to have a homogeneous composite, and the latter was diluted in sterile soil (121°C, 2 h) in proportion (1, 1/4, 1/16, 1/64, 1/256, 1/1,024) with 5 repetitions for each dilution level. In parallel, the control soil was diluted in the same way. Maize was used as an endophytic plant to trap the native mycorrhizal complex associated with A. cyanophylla. Corn seeds were pre-germinated in Petri dishes at an ambient temperature of 25°C and then planted in plastic cups containing 100 g of dilution soil. After 4 weeks of cultivation, the plants were harvested and their root systems were carefully prepared for staining. During reading, roots representing at least one point of infection were classified as positive and negative ones were presented by lack of colonization (Ouahmane et al., 2007a; Ramos-Zapata et al., 2010). The estimation of the MPN was performed using the table of Fisher and Yates (1970) and the following formula: Log  10  MPN = ( x ⁡ log ⁡ a − K ) (5) where x is the average number of mycorrhized pots, a is the dilution factor and k is a constant from table VIIItbl of Fisher and Yates for fivefold dilutions.

The standard deviation of log MPN is estimated and given by the formula (Cochran, 1950). σ  log ⁡ 10  MPN = 0.55 log ⁡ a n (6) where n is the number of samples per dilution and a is the dilution ratio.

The evaluation of modulation of A. cyanophylla shrubs was carried out by direct counting of the functional nodules formed on the root system encountered in 1-kg rhizosphere soil.

The results were processed with XL Stat software and SPSS software using the analysis of variance test (ANOVA) at a threshold of statistical significance set at 5% and the Tukey’s HSD test for the comparison of the means.

The soil pH and electric conductivity analysis showed significant differences between the rhizosphere soil and the control (bare soil) (Table 1). The introduction and establishment of golden wreath wattle (A. cyanophylla) in the sand dune of Essaouira resulted in widespread acidification of the soil in the study area. The reading of pH showed that the rhizospheric soil samples had values that were 7.4–8.6% lower than those of bare soil samples, which range from 9.10 to 8.3–8.4. Similarly, the presence of A. cyanophylla had shown a significant influence on electrical conductivity, which raised by 14.6–17.1% in rhizospheric soil (sites 1, 2 and 3) compared to bare soil (Table 1).

A significant increase in the carbon and consequently in the organic matter contents of the sampled soils in the rhizosphere of A. cyanophylla was recorded. In comparison to adjacent bare soil, the amount of carbon in the rhizosphere rose approximately more than twice in the presence of trees at all sample locations. The pleading restitution of carbon to soil is certainly linked to the massive production of foliage by this legume (Table 1). The analysis of the rhizospheric soils under the crown of A. cyanophylla shrubs at the three sites studied showed organic matter contents two to three times greater than those in the bare soil. Mineral analysis of the soils showed a substantial accumulation of the main nutrients in the rhizosphere of A. cyanophylla. A significant difference was recorded for all the analyzed elements (N, P, K ⁺ , Ca²⁺ , and Na ⁺ ) compared with non-rhizosphere soil without the influence of the shrub. Enrichment of the soil by the golden wreath wattle was noticed for all the analyzed elements and which testified their contents doubling (P, K ⁺ , Ca²⁺ , and Na ⁺ ), particularly the nitrogen contents were three (site 1) to four (sites 2 and 3) times higher in rhizospheric soil than in control soil (Table 1). The quantity of available phosphorus and total potassium in the rhizosphere enhanced roughly 1.7–1.8 times in the presence of tree roots when compared to bare soil at all study sites as compared to bare soil.

Extraction and counting of spores from subjacent soils of A. cyanophylla showed at least 800 spores per 100 g of rhizosphere soil, while non-rhizosphere soil showed 160 spores per 100 g of soil only (Figure 1). A cyanophylla had shown a greater capacity to establish a symbiosis with arbuscular mycorrhizal fungi and higher production of fungal spores in the soil. The community of AMF associated with A. cyanophylla was described by the relative abundance of different morphotypes of encountered spores. Observation of the spores under the binocular microscope at high magnification revealed the presence of four different morphotypes depending on their color: black, dark brown, light brown, yellow and grey (Figure 2). As demonstrated in Figure 3, each AMF morphotype is represented by its relative abundance in Acacia cyanophylla rhizospheric soils as well as in control (bare soil).

Morphotype1: dark brown spores; this is the most common group in different plants-sites with a relative abundance of 40–50%.

Morphotype 2: light brown spores which exhibit a relative abundance of 30–47%.

Morphotype 3: black spores with a relative abundance of 15–23%.

Morphotype 4: greyish transparent spores with a relative abundance of only 3–7%.

Microscopic observation of the roots of A. cyanophylla in the various sites showed that all the plants presented an arbuscular mycorrhizal fungus infection, and at least the mycorrhizal frequency reached 90%. The mycorrhizal colonization assessment showed rates between 40 and 50% (Table 2). The intensity of mycorrhizal infection (Ma %) varied between 40, 44, and 50% of roots collected in rhizospheric soil at sites 3, 1, and 2, respectively. It was discovered that the levels of vesicles and hyphae varied according to the location of the plants and rhizospheric soil collected; for example, Acacia cyanophylla roots from site 1 had the highest proportion of vesicles (60%), while those from sites 2 and 3 had the highest percentage of hyphae (70%). The proportion of arbuscules, on the other hand, is almost identical (10%) throughout the three sampling sites. A. cyanophylla roots showed higher levels of colonization by arbuscular mycorrhizal fungi, indicating that A. cyanophylla is one of the symbiotic plants that weave a strong relationship with AMF in their natural habitats.

Maize plants have been used as an endophytic plant to trap the native mycorrhizal complex naturally associated with A. cyanophylla. Corn seedlings were grown on different dilutions of rhizosphere soils (1; 1/4; 1/16; 1/64; 1/256; 1/1,024) from different sites under A. cyanopylla. A seedling is considered to be mycorrhized when it has at least one point of infection. Only one among five corn seedlings was mycorrhizal in the control soil, which confirms the lower number of spores in this soil (Figure 1). However, all the roots of Maize seedlings grown in rhizospheric soils of A. cyanophylla exhibited a mycorrhizal infection with different colonization rates. The number of fungal propagules (viable spores, vesicles, root fragments) in each soil was determined using the table of Fisher and Yates. Based on the calculation, the MPN per 100 g was: MPN (IC 95%) = 1,240 for rhizosphere soil and was MPN (95% CI) = 48 for non-rhizospheric soil (Figure 4). The mycorrhizal fungal community of A. cyanophylla rhizosphere soil was assessed for its self-regeneration by testing the most probable number of fungal propagules in the soil without distinguishing between these propagules (viable spores, vesicles, root fragments).

The direct counting of the total number of nodules formed by symbiotic rhizobacteria on A. cyanophylla roots showed that all the rhizosphere soil samples contained nodulated root fragments. All the root fragments at the level of the rhizosphere soil samples were examined and approximately 130 nodules per kilogram of soil were noted (Table 2).