Introduction 

Around 80 million hectares of Iran are occupied by arid and semi-arid lands, with five million hectares as active sand dunes ([2]). Therefore, large de-desertification programmes have become necessary. One method to fix sand dunes consists in afforestation ([21]). This action not only increased plant coverage and soil nitrogen (N), but also may sequester atmospheric carbon (C) dioxide ([14], [22]). Moreover, vegetation cover of sand dunes has been reported to increase soil moisture ([20]) and soil nutrients ([30]) and to restore plant diversity ([4]).

Despite the fact that sand dunes afforestation could result in high ecosystem C stock ([11]), not all plant species can survive in such a harsh environment ([19]). In fact, water shortage creates a serious limitation for successful plantations ([15]) even though other factors may also play an important role, e.g., plant species ([40]), stand age ([34]), and management ([17]). Prosopis juliflora is known as a species with good performance in arid and semi-arid areas for afforestation ([36]) and this species has been planted in several parts of Iran. However, the effects of sand dunes afforestation with P. juliflora on soil C and nutrients remain still poorly understood, though some preliminary information about the impacts on soil physico-chemical properties have been previously reported ([18]).

The objectives of this study were to determine the effect of P. juliflora sand dunes afforestation on soil C stocks and the nutrient status of afforested sand dunes. In addition, we determined the variation of soil C stock and soil physico-chemical properties under different canopy coverage classes. We hypothesized that increasing the canopy coverage could result in higher soil C stock and nutrient status of sand dunes.

  Material and methods 

Study site

The study site was located towards the south of Susa and north of Ahvaz at latitude 48° 17′ 16″ N and longitude 31° 37′ 60″ E in Khuzestan province of Iran. This site suffers from active sand dunes mostly covered by Cyperus rotundus, Pennisetum orientallis, Echinops kermanshahanicus and Artemisia vulgaris. The mean annual temperature and mean precipitation at the site are 23 °C and 251 mm, respectively (Fig. 1); soil can be classified as an Entisol; slopes are between 0 and 2% and altitude between 25 and 49 m a.s.l. The site studied was covered by petroleum mulch in 1991 and an afforestation program with Prosopis juliflora started in 1993.

Fig. 1 - Location of the study site and different canopy coverage classes (A: canopy coverage of 25-50%; B: More than 75% canopy coverage; C: control dunes).

  Enlarge/Shrink   Download   Full Width  Open in Viewer

Soil sampling, analysis and carbon stock

We selected two afforested sand dunes (10 ha each) with two levels of canopy cover (25-50% and >75%) and a nearby non-afforested sand dune (10 ha) as control. Canopy coverage was estimated using a dot grid on aerial photography ([7]). At each site fifty 20 × 20 m quadrat plots were randomly selected ([3]) and five soil samples from four corners and the centre of each plot were taken in April 2015 using a manual auger. The five soil sub-samples were pooled by depth (0-5 and 5-50 cm) and plot, and then transferred to the laboratory for the further analyses. The soil C stock (Cs, tC ha^-1) was computed using the following equation ([38] - eqn. 1):

where Bd is the soil bulk density (g cm^-3), d is the depth of sampling (cm) and C is the soil organic C content (%).

Soil organic C and total N were respectively determined by the methods of Walkley & Black ([39]) and Bremner & Mulvaney ([5]). Soil potassium (K) and available phosphorus (P) were extracted using ammonium acetate ([27]) and Olsen et al. ([31]) method. The soil texture was defined using the hydrometric method ([13]). Soil pH and electrical conductivity (EC) in deionized water were measured using the method proposed by McLean ([25]).

Statistical analysis

The data were checked for normality and homogeneity of variance using the Kolmogorov-Smirnov test and the Levene’s test, respectively. Differences in soil C stock and soil physico-chemical properties between afforested sites and control were tested using a two-way ANOVA followed by LSD test. In addition, Pearson’s correlation analysis was used to detect relationships between soil C stock and physico-chemical properties. All analysis were performed using the software SPSS^® ver. 15 (IBM, Armonk, NY, USA) for Windows^®.

  Results 

Soil physico-chemical properties

Our result indicated that the interactions between soil depth and canopy classes are significant for soil C stock and physico-chemical properties. The soil texture at a depth of 0-5 cm in the afforestation and control site were respectively sandy-loam and sandy. The soil texture at 5-50 cm was sandy in both the afforestation and control sites. A significant increase of 44 % was observed in topsoil (0-5 cm) silt percentage in afforestation compared to control sites. No significant differences in silt percentage were observed between sites at 5-50 cm depth (Tab. 1). Soil EC was significantly affected by both canopy cover classes and soil depth. The lowest EC value belonged to control dunes, followed by 25-50 % canopy coverage and next 75 % canopy coverage (Tab. 1). Soil pH was significantly decreased in afforestation sites compared to the control dune (Tab. 1). In the afforestation sites, soil P was significantly higher at both soil depths compared to the control dune. The lowest soil K was observed in the control dunes for both depths as afforestation resulted in a significant increase in soil K (Tab. 1). Although the soil N values in studied sites were low, afforestation led to significant soil N increase in topsoil (0-5 cm) compared to the control dunes, with >75% canopy coverage sites having the highest N levels in topsoil. In deeper soil depth, no significant difference was observed between afforestation and control dunes (Tab. 1).

Soil C stock in afforested and control dunes

Topsoil C stock (SOC) showed an increasing trend from 0.54 to 4.50 tC ha^-1 in control to afforestation dunes (Tab. 2). Despite the significantly higher carbon stock in afforested areas, no significant differences were observed between the two canopy classes (Tab. 2). The same SOC patterns related to afforestation were observed at the deeper soil depth and no significant difference was observed between the soil depths studied in afforested dunes (Tab. 2).

The lowest C/N ratio belonged to the subsoil (5-50 cm) in control dunes, and the highest value belonged to topsoil in the 25-50 % canopy class (Tab. 2). Furthermore, the C/N ratio decreased with soil depth in both afforestation and control sites (Tab. 2).

The soil pH and sand content in topsoil showed significant negative correlations with C stock, while positive correlations between soil C stock and all other studied parameters were observed. (Tab. 3).

Considering the significant correlations between soil silt and the rest of soil parameters, C/N and silt were important in determining soil physico-chemical properties of sand dunes in topsoil (Tab. 3).

Unlike the topsoil, soil C stock in subsoil significantly correlated with only soil P, EC and C/N ratio. Additionally, C/N ratio was positively and significantly correlated with soil EC and silt and negatively with soil N and sand (Tab. 4).

  Discussion 

Afforestation with P. juliflora increased soil C content to a depth 50 cm compared to control dunes. This is consistent with predictions for nitrogen fixing trees like P. juliflora ([33]), and with the results of Hu et al. ([14]).

Although canopy coverage class did not have any significant effect on soil C stock, our results showed a slight increase in soil C stock with higher canopy coverage. This pattern is consistent with Schulp et al. ([35]) who stated the importance of overstory on soil carbon input. Overstory can change light penetration and moisture content ([6]), which can in turn affect soil C stock ([10]). Additionally, tree spacing and canopy closure influence temperature, moisture, litter, and consequently change decomposition rate and soil C stock ([32]).

As in Garcia-Franco et al. ([11]), afforestation in the current study resulted in a higher soil C/N ratio. Changes in soil N with change in land use from sand dunes to afforested area was also seen by Li et al. ([23]). Increased soil N in the sand dunes after afforestation is likely responsible for increasing soil carbon sequestration ([12]). Higher C/N ratio and also C stock at the topsoil in sand dunes indicated the importance of this thin soil layer in C sequestration of arid lands ([28]).

Canopy coverage is known as an important factor influencing topsoil nutrients ([16]). Significant effects of canopy coverage and litter input on soil N and C have been previously observed in topsoil ([8]). This may be due to the higher litter fall in higher canopy coverage class ([1]). Similar results were observed in EC and pH, indicating the significance of canopy coverage and afforestation on the parameters stated, which is consistent with the study of Naseri et al. ([29]). Likewise, decrease in EC with soil depth was seen by Li et al. ([21]).

Soil K and P were previously shown to be significantly affected by afforestation, with an increasing trend from bare to afforested dunes ([24]). This might be related to the soil pH reduction providing better conditions for these two nutrients availability in soil ([26]). Soil texture has been reported among the factors strongly affecting the soil carbon dynamic ([32]). In our study, the soil silt and sand respectively had positive and negative correlations with soil C stock, which could be due to longer residence time of soil organic carbon in silt ([9]). Improvement of soil physico-chemical properties and soil C stock after two decades of afforestation demonstrates the remarkable importance of P. juliflora in the parameters studied and long term fixation of sand dunes.

Soil fertility and organic carbon were important factors on C stock in arid lands, as soil C stock and physico-chemical properties highly correlated ([11]). However, afforestation resulted into soil pH decreasing from alkaline toward neutral. Soil pH negatively affected on the C stock. Considering these two, we could conclude that soil pH declines after afforestation and prevents carbon loss to the atmosphere ([37]).

  Conclusion 

P. juliflora afforestation in sand dunes resulted in higher soil C stock. Higher C stock in topsoil and high level of correlation between soil C stock and almost all soil physico-chemical properties studied, especially in topsoil, indicated the importance of soil features in soil C stock of arid lands. Furthermore, canopy coverage is considered a factor affecting soil N, which consequently resulted in higher soil C stocks in sand dunes of Iran.

  Acknowledgments 

We thank Behbahan Khatam, Alanbia University of Technology, Iran, for financial supporting this research work. Special thanks to Dr. David Showalter for his kind help in improving the English level of this work.

  References