The Effects of Mesquite (Prosopis juliflora) on Soils and Plant Communities in the Deserted Rangelands of Bahrain

The influence of mesquite trees (Prosopis juliflora (Swartz) D.C.) on the physicochemical properties of soils and annual understory plants was investigated in the deserted rangelands of Bahrain. Soil properties were measured in the understory and the uncanopied adjacent areas of mesquite trees. Likewise, the number of plant species was assessed in four 1 × 1 m randomly distributed quadrates in the understory and the uncanopied adjacent areas. The results showed that sand particles exceed 96% of soil composition. Soil bulk density at the 0–5 cm soil depth was significantly higher in the understory of trees compared to the uncanopied adjacent areas. However, moisture at a depth of 40–60 cm was significantly higher in the uncanopied adjacent areas. No differences in the pH, EC, K, Na, and Ca were found between understory and the uncanopied areas in all soil depths. Levels of N, P, Mg, and organic matter were significantly higher in the understory of trees compared to the uncanopied adjacent areas. Organic matter was twice the amount in the upper 20 cm of soil layers in the understory of mesquite trees. Species richness did not differ between the understory and the uncanopied areas. Nevertheless, the density of ephemerals in the understory of mesquite trees was higher than the uncanopied areas by 18%. The Shannon-Weaver index of diversity was higher in the uncanopied areas compared to the understory. The study concluded that the canopy effects of mesquite trees on soil vary with depth. Nonetheless, the influence of mesquite on flora could be beneficial for annual understory plants but subject to many operating factors, including density and cover of mesquite trees.


Introduction
Worldwide, invasive alien species are considered as direct drivers of ecological change and biodiversity decline [1,2]. ese species cause damage to the environment and human health.
ey function as agents of change that influence other species, habitats, and ecosystems [3]. Prosopis juliflora (Swartz) D.C., commonly known as mesquite, is one of the species that is reported to be invasive in many regions of the world [4]. It is an evergreen tree belonging to the Leguminosae family. Mesquite is a nitrogen-fixer species [5] and thrives in low rainfall and high-temperature areas. It grows on compacted soils or soils with hardpans, tolerates saline soil, and efficiently extracts soil water [6,7]. e species was introduced into Bahrain for landscaping and amenity purposes in the 1930s. e oldest specimen of the species, which is believed to be 500 years old, locally known as the "Tree of Life," is considered a focal point in the landscape of Bahrain [8]. Mesquite trees are naturalized and widely planted in towns and villages of Bahrain. Recently, mesquite trees have been sighted along drainage canals, agricultural farms, and deserted rangelands as well as abandoned farms in the country.
Several authors studied the influence of mesquite trees on soil properties [9][10][11][12]. Nonetheless, such an effect was linked to other relevant factors, including climate, soil properties, canopy density, and plant coverage [13]. For instance, Tiedemann and Klemmedson [10] and Deans et al. [14] stated that surface soils under crowns of Prosopis spp. were more fertile than soils in the adjacent areas. Likewise, several scholars [15][16][17] highlighted the positive effects of mesquite on organic matter and nutrient contents of the soils.
On the other hand, adverse effects of mesquite on biodiversity and ecosystems including rangelands, pastures, wetlands, and native flora have been controversial [4,[18][19][20][21]. In this respect, Mukherjee et al. [22] reported that mesquite is demising dominant native species over a long time. El-Keblawy and Al-Rawai [16] and El-Keblawy and Abdelfatah [19] stated that associated plant communities were negatively affected under the canopies of mesquite trees, and the magnitude of the effect was dependent on the size and density of the trees. On the contrary, Naudiyal et al. [23] indicated that P. juliflora canopy had a little effect on local flora, and mesquite facilitated the regeneration of native species under its canopy. e effect of mesquite trees on soil properties and understory plants has been altered by many parameters including the climate of the region [24], vegetation types [25], stand attributes [16], and soil properties [17]. Conceivably, mesquite invasion and its threat mechanism to different habitats and flora in Bahrain are neither known nor documented. ere is a need to understand the process of mesquite invasion, the extent of its colonization, impacts, interactions, and analysis of the risk associated with its spread. e objective of this study is to determine the influence of P. juliflora on soil properties and annual understory plants in the sandy ecosystems of Bahrain.

Study Area.
Bahrain is an archipelago of 36 islands occupying a total land area of 779.39 km 2 . It is located between latitude 25°32′ and 26°20′ north and longitude 50°20′ and 50°50′ east. e study area is chosen to be in the middle of one of the largest mesquite stands in the country at 26°10′20″N and 50°29′36″E ( Figure 1). e environment is arid to extremely arid with mean annual temperature and precipitation of 26.8°C and 99.8 mm, respectively. Rainfall is erratic and occurs in winter only while aridity (P < 2T) prevails all year long [26] (Figure 2). e mean daily evaporation rate ranges from 3.5 mm in January to 15.6 mm in June. e soils are shallow, poorly defined in the structure, and mostly underlain by sand or a cemented layer of the gypsic or calcic horizon [27]. Major soil groups include Solonchaks, Regosols, Yermosols, and Fluvisols [28].

Soil Analysis.
Soil samples were collected from the understory (2 m from the main stem of the tree) of four randomly distributed mesquite trees and the uncanopied adjacent areas (2 m distance from the edge of the tree crown). A core was taken at four depths of 0-5, 5-20, 20-40, and 40-60 cm in four replicates. e samples were air-dried and passed through a 2 mm sieve. Bulk density (BD) (g·cm −3 ) and soil moisture content (SMC) were estimated [29]. e texture was determined using the hydrometer method. e soil was assigned to a textural class according to the USDA textural triangle. Electrical conductivity (EC) and pH were measured in a 1 : 1 (soil : water) extract using Orion Star ™ A222 conductivity portable meter. Soil organic matter (SOM) was estimated using the mass loss on ignition method by placing samples in a muffle furnace at 440°C overnight. Nitrogen percentage was determined using the micro-Kjeldahl method. Nutrients (P, K, Na, Ca, and Mg) were processed according to EPA 3050B method and extracts measured using inductively coupled plasma optical emission spectroscopy (ICP-OES). Means were analyzed at P < 0.05 significance level in two-way analyses of variance using the software statistical package JMP 14 [30].

Flora Analysis.
e heights and diameters of mesquite trees were measured in four randomly quadrates of 50 × 50 m. Canopy cover was estimated using a 100 m line intercept running along the middle of each quadrat [31]. e crown diameter of trees canopying the small quadrates of annual species was measured in two perpendicular directions. Annual species were measured in a four quadrate of 1 × 1 m established under the understory and the uncanopied areas of four randomly selected mesquite trees. Within the quadrates, existing annual plants were tallied. Species coverage was assessed using the Daubenmire method plot [32]. e relative density and frequency of species were calculated [31]. Shannon-Wiener index of annual species biodiversity (H′) was calculated [33].

Soil Physicochemical Properties.
e results of the study indicated that sand comprised 96.4% of soil composition, while clay and silt constituted 3.4% and 0.2%, respectively. Soil BD ranged from 0.92 to 1.55 g·cm −3 . An increasing trend in BD with depth was detected in the understory and the uncanopied adjacent areas. Significant differences in BD were observed between the understory of mesquite trees and the uncanopied adjacent areas at a depth of 0-5 cm (Figure 3). e low BD of soil in the understory could be attributed to the deposition of aeolian materials in the uncanopied areas. In this respect, our results are in contrast with Tiedemann and Klemmedson [10], who reported lower bulk density in the 0-4.5 cm soil depth under mesquite trees compared to the open areas in southern rangelands of Arizona.
Moisture content in the soil profile ranged from 3.41% to 5.26% in the understory compared to 3.10%-21.9% in the uncanopied adjacent areas. No significant differences were observed in SMC among various depths of the soil profile in the understory and the uncanopied adjacent areas, except for the 40-60 cm layer (Figure 4). e observed lower soil moisture in the 40-60 cm layer in the understory of mesquite trees could be attributed to water root uptake from the lower level of the soil profile. Similar results were reported by Schade et al. [34], as indicated that P. velutina could use up soil water.
Soil pH was moderately alkaline, ranging from 7.9 to 8.1. e maximum pH was 8.1 ± 0.03 in 0-5 cm depth in the uncanopied areas compared to 7.9 ± 0.04 in the 20-40 cm depth understory. No significant differences in pH were observed between understory and the uncanopied areas. Electrical conductivity ranged from 1.44 to 3.56 mS/cm. Likewise, no significant differences in EC were found beneath and beyond the canopies of mesquite trees (Table 1).
Significant differences existed in SOM in 0-5 cm and 5-20 cm depths between the understory and the uncanopied areas. e SOM was more than twice the amount in the first two topsoil layers than those of uncanopied adjacent areas. However, a decrease in SOM was noticeable from the topsoil downwards in both understory and the uncanopied adjacent areas. e low level of SOM in this study compared to other ecosystems could be attributed to variations in moisture, pH, vegetation cover, land management, and land-use history [35][36][37]. Some scholars reported that mesquite trees could produce up to 6.1 tons per hectares per year [38]. In our case, harsh climatic conditions and sparse vegetation cover were the main factors contributing to low SOM content.
Likewise, significant differences existed in N percentage between the understory and the uncanopied areas, down to a depth of 20 cm. Nitrogen was twice the amount in the 0-5 cm and 5-20 cm depths in the understory compared to uncanopied adjacent areas. Our results confirmed those of Tiedemann and Klemmedson [10], Menezes and Salcedo [11], El-Keblawy and Al-Rawai [16], Mosweu et al. [39], and Najafi and Jalili [24] in different regions of the world. Many authors reported high nitrogen levels underneath trees and attributed that to the accumulation of organic matter within the crown boundaries and the nitrogen-fixing ability of the plants [40,41].
A similar trend was found for the phosphorus, where concentration was higher in the depths of 0-5 cm and 5-20 cm in the understory than the comparable depths of uncanopied adjacent areas. Equally, significant differences existed in Magnesium (Mg) between the understory of mesquite trees and the uncanopied adjacent areas, starting from topsoil to 20-40 cm depth. e result is consistent with Tiedemann and Klemmedson [10], Menezes and Salcedo [11], El-Keblawy and Al-Rawai [16], and Mosweu et al. [39]. e concentration of K showed a decreasing trend with soil depth in the understory, but not significant, compared to the uncanopied adjacent areas. e maximum concentration of K was 190 ± 122 in 0 to 5 cm in the understory of mesquite trees. However, the minimum level of K was 50.6 ± 8.06 in 20 to 40 cm depth in uncanopied areas. e result attested to the findings of Tiedemann and Klemmedson [10] and Menezes and Salcedo [11]. e decrease in K suggests that trees of P. Juliflora absorb K from deeper layers and accumulate it on the top of the soil profile.
No significant differences in the levels of Na and Ca were observed between the understory and the uncanopied areas. Furthermore, no clear trend in these nutrients was detected with soil depth. Comparable results were reported by El-Keblawy and Al-Rawai [16] in the UAE desert and Kahi et al. [42] in the rangelands of Kenya as well as Menezes and Salcedo [11] in northeastern Brazil. e high concentration of some nutrients in the topsoil of P. juliflora understory was attributed to its high nutrient demands that impoverish lower soil profile and release nutrients when leaves decay [43].

Flora Impact.
e results of the study showed that mesquite tree heights and diameters averaged 7 ± 1.57 m and 20 ± 8 cm, respectively. Trees were of a shrubby multistemmed form ( Figure 5). e number of stems ranged between 5 and 6. e crown of trees ranged from 3.8 m to 27.8 m in diameter, averaging 12.5 ± 6.8 m. Large openings (up to 20%) existed in old trees due to the dieback of branches. e density of mesquite trees was 21 ± 1.31 treesha −1 , while the canopy covers averaged 14.68%. Associated flora was of an ephemeral type with a total of seven identified species ( Table 2). Some of them like Aizoon canariense L. and Emex spinosa (L.) Campd. are of considerable antioxidant activities [44]. Species richness (n � 6) was the same in the understory of mesquite trees and uncanopied adjacent areas. However, the density of  International Journal of Forestry Research ephemerals in the uncanopied areas was higher than the canopied by 18%. e presence of forbs in the understory of mesquite trees was attributed to light shade created by openings in the canopy of trees [42]. Tiedemann and Klemmedson [10] attributed abundance and good growth of grasses under mesquite crowns to improved soil conditions, resulting from the concentration of nutrients from the lower layers. e Shannon-Weaver index of diversity was higher in the uncanopied areas compared to the understory, calculating 1.27 and 1.18, respectively. Species evenness (Shannon Equitability (H/ln N)) index in the uncanopied areas was 70.7% compared to 65.6% in the understory. e results of this study were in line with El-Keblawy and Al-Rawai [16], who reported lower frequency, density, and diversity indices of native species underneath canopies of mesquite trees in the rangeland of UAE.
Finally, since its introduction into the country, mesquite widely spreads in the rangelands of Bahrain due to seeds dissemination by camels. Recently, mesquite was categorized as slightly invasive with low invasion risks [45]. Nonetheless, mesquite thrives a harsh environment and provides shade for native flora and fauna. erefore, mesquite offers valuable environmental services in desert ecosystems through halting desertification and encountering the climate change effects in desert countries like Bahrain.

Conclusion
is study presented the influence of mesquite as an invasive species on some physicochemical properties of sandy soils and annual plants in Bahrain. e study showed mixedeffects of mesquite trees on soil's properties. No differences in the pH, EC, K, Na, and Ca were found between the understory and uncanopied adjacent areas. However, significant differences existed in soil organic matter, N, P, and Mg between the two categories in the upper 20 cm of the soil. Mesquite trees did not affect species richness; nevertheless, species diversity was higher in the uncanopied areas compared to the understory.
Based on the results of this study, it can be concluded that mesquite enriches upper soil layers with some nutrients and organic matter. Moreover, its adverse effect on annual plants in the arid environment where sandy soils prevail is minimal. Mesquite proved to be an invasive and highly competitive species in many ecosystems of the World. Nonetheless, it did not function as an invasive species in the low-resources sandy ecosystem of Bahrain. Mesquite tolerates drought and fixes nitrogen in the soil. It stabilizes sands and nurses some annual understory plants. erefore, we recommend planting the species to reclaim similar sandy ecosystems in the arid environment. In the meantime, we believe that critical knowledge gaps still exist regarding the

Data Availability
Data are available upon request from the corresponding author.

Disclosure
Ahmed A. Salih was formerly affiliated with Natural Resources & Environment Department, Arabian Gulf University. Jameel A. Alkhuzai was formerly affiliated with the University of Bahrain.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this work.