Introduction

Plant species often live in the cracks and rocky layers of arid lands (Porembski and Barthlott 2000). Rock cracks provide micro climate for arid lands vegetation and protect them against extreme environmental conditions, such as direct solar radiation (Lopez et al. 2009). Abundance of the endemic cactus, Mammillaria fraileana, is in relation with the fissure density, fracture depth of rocky layers, although these relationships show weak significant correlation (Lopez et al. 2009). Many rocky slope restoration projects carried out without proper scientific guidance may not to be very satisfactory which refer to lack information of plant growth on natural conditions of a slope (Shu et al. 2003; Wang et al. 2009). Surface crevices are the main water storage of plants in the arid rocky slopes ecosystems and are important contributors to soil quality (Powers and Ferrell 1996; Witty et al. 2003). The results of Porembski and Barthlott (2000) studies on inselbergs in arid zones indicated the role of granite rocks to support vascular plants against water limitation. In other hand, the role of succulent plants on breakdown igneous and degradation of volcanic rocks has been studied in desert area (Bashan et al. 2002, 2006). Other studies in humid and semi-arid climate areas have shown relation of heterogeneity and spatial patterns of plants with limestone dispersions and Karst development (Benites et al. 2007; Pérez-García and Meave 2004; Zhang et al. 2010; Lin et al. 2011; Zwieniecki and Newton 1996). Investigations of rocky slopes at Alborz and Zagros mountains in Iran also showed growth and regeneration of woody plants on varies slopes (Salarian et al. 2008; Ravanbakhsh and Marvie Mohajer 2010). Some researchers focused on changes of groundwater aquifer and assessed that as a water harvesting system (e.g., Pazand and Sarvestani 2013; Radfar et al. 2013; Nikiema et al. 2013; Wang et al. 2013; Alvarez et al. 2012; Yangui et al. 2011; Liu et al. 2011a, b; Fernandes et al. 2010) while there are rare studies on the rocky slopes of arid lands mountains. Studies at Mehriz region, central Iran, has showed that shrubs and rock-colonizing tree species (Amygdalus scoparya, Ficus juhanica, Pistacia atlantica) are grown in fissures of limestone rocks as primary colonizers. They produced small amount of soil by their own growth or sedimentation. Therefore, these species encourage soil and water accumulation in rocky slope crevices on bedrocks (Dashtakian and Baghestani 2002).

In the Mehriz region, as a case study of this work, rainfall from late fall to spring is insignificant; therefore, trees and shrubs are heavily independent to water storage in substrate layers. The main goals of this work are to clarify how different dimensions of cracks on lithologic units contribute soil quality in arid ecosystems, and what is role of limestone bedrocks growth of trees. It was hypothesized that different measured crevice properties might be in related to local variation of vegetation cover percentage (VC %) and number of trees within different geology units. A further assumption was that crevices in different lithologic habitats would show meaningful differences. Based on Whittaker (1972) and Whittaker and Woodwell (1972) studies, geomorphic heterogeneity considered one of the most important environmental characteristics influencing the numbers of plant species in a given area for some time. In addition, high correlations of VC and tree density with crevice properties showed to emphasizing VC (%) and number of trees in different lithologic units, such as igneous and sedimentary rocks (granite, andesite and limestone).

Materials and methods

Study area position

In this research, the rocky layers of three lithologic types (limestone, granite and andesite) were studied in the Mehriz region of Yazd province, located in central Iran (54°13′52″ to 54°32′19″E and 31°22′37″ to 31°34′53″N) (Fig. 1). Based on a modified De Martonne classification method, this area has cold-arid climate (Botzan et al. 1998), with a mean annual rainfall of 189 mm and a high inter-annual variability. The mean annual temperature is 12.2 °C. In the studied region, the hottest month is August and the coldest is February, with mean monthly maximum and minimum temperatures of 24.4 and 1.8 °C, respectively. In respect to geomorphology, the studied sites are located on mountain units. Rocky slopes are the most abundant landscape types seen in these mountains (Table 1).

Fig. 1
figure 1

Location of the study area; in this map has been shown the location of granitic (blue line, image a), andesitic (yellow line, image b), and limestone (red line, image c) geology units

Full size image
Table 1 Characteristic parameters and indices of lithology for three rocky slopes of Mehriz Mountains, central Iran
Full size table

Methodology

The methodology of this study is expressed in three parts: data collection, measurement, and analysis. In this research to measure vegetation cover percentage, intercept measurements have been calculated along a transect line with 25 m length. Soil properties was also assessed along the transect line where it overlapped with soil and other deposits.

Data collection

Based on Crépin and Johnson (1993), a stratified random sampling scheme was designed for analyzing the variability within and between sites (lithologic units). 30 transects, each 25 m long, were randomly laid out throughout the study area (Table 1). Along these transects the following were measured: vegetation cover percentage (VC) and large crack width (LCW), medium crack width (MCW), small crack width (SCW), large crack depth (LCD), medium crack depth (MCD), small crack depth (SCD), large crack length (LCL), medium crack length (MCL), small crack length (SCL) and numbers of trees. Crevice numbers possessing these properties were determined with respect to the classification shown in Table 2. Classification of crevice dimensions based on their width, depth and length, was by gathering experts’ opinions and ultimately categorizing crack scales. The depths of the crevices were measured using a steel band. Positions of transects were randomly established in the landscape along vertical and horizontal gradients. Slope (%) was measured using clinometers. Aspects have been measured with respect to true north and were recorded as the azimuth (u) (McCune and Keon 2002).

Table 2 Classification of crevices according to width, depth and length (in cm)
Full size table

The stochastic methodology for transect in this research means that based on the determined altitude, transect was done for each lithologic unit under random method. In other words, although the selection of transect location was random, the distribution was logical and all of lithologic units were transected.

According to the Table 2, for each fracture dimension, a range has been determined. The goal was a classification based on the values for the width, depth and length. The volume of cracks increases with increasing width, depth or length which provides more space to keep moisture and soil to grow vegetation cover.

Soil measurements

To quantify soil characteristics, soil samples of three crevice depths ranges (0–30, 10–30 and 30–50 cm) were randomly taken on twenty crevices of each geology unit. After preparing the soil samples, they were analysed in the laboratory. Then the soil samples were passed through a 10-mesh (2 mm) screen to determine gravel content. In addition, for particles finer than 2 mm a physical test for determining clay, silt and sand amounts was applied. Soil texture was determined using the Bouyoucos hydrometer method (Bouyoucos 1962). Soil moisture contents were measured by the oven-dry method (dried at 105 °C to constant weight). Electrical conductivity (EC) was determined on a saturated soil paste. Calcium carbonate was determined by neutralization of HCl and titration of NaOH (Nelson 1982). Soil acidity (pH) was measured in distilled water by pH meter equipped with a glass electrode and calomel reference cell; total inorganic carbonate was assessed with the Colorimetric method by Spectrophotometer and the Shiebler calcimeter method (Allison and Moodie 1965). Also, total soil organic matter (SOM) was assessed with the Walkley–Black wet oxidation method (Allison 1965). Total nitrogen was determined by the Kjeldahl method (McGeehan and Naylor 1988), total potassium was estimated by flame photometry (Shi and Bao 1980) and total phosphate by the sodium carbonate (Na2CO3) and fusion––Molybdenum blue photometric method (Syers et al. 1967).

The amount of lime was calculated based on Total Neutralizing Value (TNV) through the volume method using the neutralizing reaction with HCL (Goh et al. 1993).

Data analysis

Statistical analysis of variance (ANOVA) was applied using SPSS (version 13). Tukey’s multiple-range test was used for mean separation to determine significant differences (p = 0.05) in the number of different crevice properties and soil characteristics among the three lithologic units. A Spearman’s Rank Correlation test was performed to analyze correlations between lithological properties, VC (%) and number of trees.

Results

The following, findings of the research are explained according to the properties of cracks, vegetation cover and lithologic units.

Rock crevices properties

The crevice parameter values of the three lithological types are shown in Tables 3 and 4. The cracks on the limestone with low and medium sizes in width and depth showed more significant differences compared with the same properties on the other lithologic units including andesite and granite. This difference was significant for cracks on the limestone compared with others for low length also. On the granitic rocky slopes, these parameters have higher values than those on the andesitic rocky slopes. Therefore, the values of LCW, MCW, LCD, MCD and LCL have directly significant relationships with vegetation cover, with higher crevice properties on the limestone and granitic rocky slopes, than on the andesitic rocky slope (Tables 3).

Table 3 Comparison of the crevice properties and percentage of vegetation cover among three rocky slopes in Mehriz Mountain, central Iran
Full size table
Table 4 Physical and chemical parameters of soils on the three lithology of Mehriz Mountain, central Iran
Full size table

Soil properties

Analysis of soil properties showed that only pH, gravel, lime and moisture content were significantly different among sites (Table 4). In all lithologic units, positive relationships were revealed between depth and soil moisture content, while there was a significant difference between three depths of soil within rocky layers. Limestone contains highest amount of soil moisture content that is seen in 30–50 cm depth of layers. At this depth, there was a significant difference between limestone and other lithology units with moisture content. For all crevice depths among the lithologic units, soil properties including total nitrogen, total phosphorus, total potassium, electrical conductivity, soil texture, and organic material showed no significant differences.

Spearman’s correlation coefficient between vegetation cover and crevice properties

After studying the crevice properties and soil characteristics on different lithologic units in relation to the vegetation cover of the rocky slopes and the life forms, we analyzed the relationships between these parameters by calculating the Spearman’s correlation coefficient as shown in (Table 5). A significant difference was observed in vegetation cover (VC) in the studied units. Investigations of the geology units showed that the highest percentage was in limestone (31.46 %) and the lowest vegetation density was in granite and andesite, with 10.12 and 5.93 %, respectively. In all lithologic units of the studied area, no significant positive correlation was seen among values of SCW, SCD and SCL. Also in each lithology unit involving limestone, granite and andesite, there was no significant positive correlation between the values of the mentioned properties and vegetation cover. There were significant relationships between medium width and depth cracks with vegetation cover in all lithologic units. In the limestone unit, statistically positive significant correlations were found between the vegetation density and low and medium cracks. Positive correlations were found between large and medium cracks in andesite, and vegetation cover. Also a direct correlation was found between values of low and medium dimension of cracks with vegetation cover in the granitic unit. In the limestone lithologic unit, the number of trees showed a positive correlation with the values of low and medium sizes in cracks including depth, length and width.

Table 5 Spearman’s correlation coefficient between vegetation cover and number of different crack properties in each geology unit and also combined
Full size table

Discussion

Rock slopes have a number of characteristics which inhibit root development, such as highly impervious bedrock, limited soil volume, and lack of space (Nagy and Proctor 1997; Bashan et al. 2002, 2006). Soil accumulation on rocky slopes is a key factor enabling the establishment of vegetation cover in early succession stages, but due to the rainfall impact and severe water erosion, it is difficult to maintain soil protection on a rocky slope surface (Wang et al. 2009). Sometimes, the accumulation of soil in rocky cracks provides conditions to establish plant cover (Wiegleb and Felinks 2001; Wang et al. 2009). The mineralogical composition, texture and pore characteristics of different rocks are the main factors that control the intensity of features such as cracks (Hudec 1998). This research indicated that the numbers of crack properties significantly varied with lithology (Fig. 2). There were high numbers of cracks on the rocky slopes of limestone units. Deep fractures and cracks exist between layers of limestone formation (Fig. 2) and other sedimentary rocks. This finding confirmed the opinion of Jing (2003) about the relationship between cracks and geology formations. Igneous rocky geology units (for example granitic and andesitic) have less inherent porosity and the available moisture content is poor in these geology units (Tug˘rul 2003).

Fig. 2
figure 2

Tukey mean separation for the number of crevice of the three different lithologic units

Full size image

Figure 3 shows that the moisture content at 30–50 cm depth in the limestone unit has are significantly higher than on the granitic and andesitic rocky slopes (Fig. 3). The layering structure could partially explain why there was a higher moisture content on the limestone rocky slope rather than on other lithology units. A low amount of available water is one of the most important limiting environmental conditions for plant inhabitation on the rocks (Martinez 1999; Nobel and Zutta 2007; Lopez et al. 2009; Alan 1997; Huang et al. 2005).

Fig. 3
figure 3

Tukey mean separation for the soil moisture content of the crevices of three different lithologic units [Letters of a, b, c, d, and bc show significant or non-significant differences in probability level 5 % (95 % confidence)]. bc indicates moisture regime of depth 10–30 cm in limestone between moisture regimes b and c. For instance histograms with letter of b indicate that in each of lithology with each of depth have same moisture regimes

Full size image

The reasons causing significant difference between diverse lithology units in relation to vegetation cover and life forms are as follows.

Figures 4 and 5 show how different crevice properties influence the vegetation cover in each geology unit (Figs. 4, 5). Wherever the soil is adequate on crevices and rocky depressions, there it is no doubt that the roots will be able to penetrate and anchor in the rock crevices (Li et al. 2007).

Fig. 4
figure 4

The relationship between number of LCW (a), MCD (b), SCW (c), LCD (d), MCD (e), SCD (f), LCL (g), MCL (h), SCL (i) and VC of all three different lithology units

Full size image
Fig. 5
figure 5

The relationship between number of LCW (a), MCD (b), SCW (c), LCD (d), MCD (e), SCD (f), LCL (g), MCL (h), SCL (i), and number of trees of limestone lithology unit

Full size image

Field observations showed that plant roots were mostly concentrated in bedrock crevices. Environmental factors, such as soil nutrients and soil moisture, may lead to varied patterns of root distributions (Coutts 1987; Canadell et al. 1996). The roots could be more extended in wider crevices (Li et al. 2007). Plant roots creating anchors in crevices will cause significant deformation in the crevices, especially in the areas where the crevice width is smaller than the roots sizes on the bedrock surface. Root systems in the crevices had an overall flat shape instead of a circular one when being outside the crevices (Li et al. 2007).

It is known that roots tend to grow into the crevices under drought stress and whenever the moisture content in the crevices is suitable they concentrate at higher levels. This common phenomenon on the natural rocky slopes is showed in Fig. 7. Water conditions have driven the extension of roots into the crevices (Li et al. 2007). Eight dry months between April until November indicated that considerable water stress for plant species in the studied area is a common phenomenon. Therefore, due to the different water requirement of plants and the physical variations of the growth substrate (bedrock), which are essential components of the plant water supply system and are important contributors to soil quality (Witty et al. 2003), the form of plant life will be diverse, especially in an arid region. Amigdalus scoparya is dominant tree species of limestone rocky slopes. Other tree species grown in this lithological unit are Ficus juhanica and Pistacia atlantica.

It is obvious that plant communities of rocky slopes that mainly contain tree species have relatively few desiccation-tolerant species (Porembski and Barthlott 2000). Therefore, under the same climate, these plant communities were only established on the lithologic units with more available water in the surface crevices. The ability to take up stored water in the surfaces of crevices on the bedrock during the pronounced dry season is likely the key feature allowing plant species to thrive under dry conditions in the rocky slopes of the study area (Querejeta et al. 2006). The limestone lithological unit had the largest amounts of crevice parameters and vegetation covers, while only on this unit were tree species established. Tree species almost exclusively developed over limestone lithology rocky slopes. Plant life form of developed vegetation cover on the granitic and andesitic rocky slopes was only shrub species (Fig. 7). The main role of bedrock involves supporting tree transpiration during dry periods in ecosystems with shallow soils (e.g. Zwieniecki and Newton 1996; Hubbert et al. 2001; Rose et al. 2003). Furthermore, subsurface soil-filled cavities within the limestone bedrock can hold large volumes of water which can better prevent evaporation compared with water stored in superficial soil (Querejeta et al. 2007).Water stress was one of the most important limitative factors to grow species on rocky slope (Wang et al. 2009).

Plant species needing low water requirement such as shrubs are mostly established in small and medium size crevices, but those having a high water requirement, such as trees, are mostly established in large crevices (Figs. 6, 7a). Therefore, it was clarified why seedlings of plant species (especially tree species) in the study site have frequently extended their roots into the surface crevices (1–4 m) of different lithology units (Jackson et al. 1999; Stratton et al. 2000; Sternberg et al. 2002), but only low numbers of those survived during the pronounced dry seasons.

Fig. 6
figure 6

Established shrub species on the small width and deep crevice of the andesitic rocks (a), vegetation associated with granite boulders and crevices of granitic rock (c), Ficus juhanica tree growing in a limestone crevice (e). Ficus juhanica tree growing in crevice located between two layers (upper layer and lower layer) on a limestone rocky slope (e), natural restoration of Amygdalus scoparya and Pistacia atlantica in a crevice on a limestone rocky slope (d and b, respectively). Schematic of limestone rocky slope layering (g). Surface crevice helped shrub species to establish on the limestone rock (h)

Full size image
Fig. 7
figure 7

Schematic of growth substrate for plant species on the three rocky slopes. a Limestone, b andesite and c granite

Full size image

Due to the direct relationship between soil moisture content and increase of crevice depth as well as the significant difference between the number of crevice parameters in limestone, granitic and andesitic lithology units, the results explain why there were significant differences between the three different lithological units in respect to vegetation cover and life forms (Lopez et al. 2009). In summary, this is probably the first or one of the first surveys to establish the way in which different lithological units affect the inherent potential for plant establishment on arid lands rocky slopes.

Conclusion

Lithological deserts and rocky landscape occur over wide parts of arid areas in Iran. Establishment of vegetation cover in these severe landscapes occurs in places where competition for water is low. Crevices and cracks on the rocks are locations for supplying water to establish trees and vegetation cover in these areas. The objective of this study was to provide an overview about the important role of surface crevice properties of different lithologic units in influencing vegetation cover and plant life forms due to plant water supply and soil accumulation. Investigations of this research indicated that limestone units because of their greater ability to keep water in their cracks are more preferential sites for trees and shrubs. Also there is a significant correlation between the width size of cracks with life forms as well as for length and depth sizes. Igneous rocky geology units (for example granitic and andesitic) have less inherent porosity and the available moisture content is poor in these geology units. Results indicated that limestone unit with more than fractures and cracks compared granite and andesite units, are more suitable locations to establish vegetations. The survey results can be utilized for forest restoration projects in arid land mountains.