Geoderma 2020 Omanterraceagriculturefinal

Geoderma 363 (2020) 114152
Contents lists available at ScienceDirect
Geoderma
journal homepage: www.elsevier.com/locate/geoderma
Terrace agriculture in a mountainous arid environment – A study of soil T

quality and regolith provenance: Jabal Akhdar (Oman)
Daniel Moraetisa, , Sumaya Salim Al Kindib, Sara Kalifah Al Saadib,
⁎
Ahmed Abdul Raoof Ali Al Shaibanic, Kosmas Pavlopoulosd, Andreas Scharfb, Frank Matternb,
Michael J. Harrowere, Bernhard Pracejusb
a
Department of Applied Physics and Astronomy, University of Sharjah, P.O. 27272, Sharjah, United Arab Emirates
b
Department of Earth Sciences, Sultan Qaboos University, Al-Khod, Muscat 123, Oman
c
Petroleum and Chemical Engineering Department, Sultan Qaboos University, Al-Khod, Muscat 123, Oman
d
Sorbonne University Abu Dhabi, Geography and Planning Department, P.O. 38044, Abu Dhabi, United Arab Emirates
e
Department of Near Eastern Studies, Johns Hopkins University, Baltimore, MD 21218, USA
ARTICLE INFO ABSTRACT
Handling Editor: Dr Alberto Agnelli In the Sultanate of Oman remnants of deteriorating terrace agricultural systems offer important insights into
Keywords: long-term human adaptation in the arid tropics. Irrigation and terrace agriculture in the mountainous Jabal
Terrace agriculture Akhdar region reveal historic agricultural practices in a rugged, high elevation context. The present study ex-
Arid environments amines soil quality and regolith provenance in abandoned agricultural soil terraces. Three soil profiles in each of
Calcrete the Villages of Hadash and Wijma were excavated and analyzed. Physical, chemical and mineralogical analyses
Soil provenance were conducted for all soil horizons. In addition, six other soils, 3 possible soil parent rocks (regolith) and soil’s
bedrock were collected. Soil ages were constrained by 14C assays and stable isotope, (13C and 18O) on the bulk
carbonates in the calcrete (caliche). The results demonstrate that both sites display poor soil quality with very
low average total organic carbon (TOC) (6.2–5.0 g kg−1) and mean weight diameter (MWD; 0.27–0.48 mm),
with low water stable aggregate content (< 42%). All the geochemical, mineralogical and the thin section
analyses show that the soils exhibit unique characteristics that differ from those of other sediments (possible
parent regolith) and soils in the vicinity. The finding of ostracod shells in the soil terraces in both areas and 14C
dating of calcrete (10.193 ± 30–13.887 ± 40 a BP) indicate that regolith was human-transported to terraces
to create soil. The 14C ages of the bulk carbonates match well with a dry period of high calcite precipitation
contemporaneous to the Younger Dryas. The Hadash and Wijma soil terraces are located ~45 km away from
each other, but still display significant similarities in terms of regolith provenance and soil development and
were likely filled with regolith from the same source. These results offer new perspective on agricultural terrace
development and oasis agriculture in a rugged, high-elevation, arid environment.
1. Introduction positive or negative influences on soil fertility and erosion (Kosmas

et al., 2000b; Novák et al., 2014). There are several studies on soil
Soil terraces are a mode of agriculture found in many regions abandonment resulting in improved soil quality under Mediterranean
worldwide (Varotto et al., 2019). Constructing soil terraces to create climatic conditions and with specific site conditions of slope and soil
arable land in rugged terrain is one of the most labor-intensive forms of lithology (Dunjó et al., 2003; Kosmas et al., 2000a,b). On the contrary,
agriculture practice in arid regions. In semi-arid circum-Mediterranean abandonment of soil terraces in more arid environments can be detri-
areas, soil terraces are still a common form of agriculture (Moraetis mental to soil quality given limited precipitation and the risk of mineral
et al., 2015; Tarolli et al., 2014). However, large areas of terraced precipitate accumulation. In Yemen, soil terraces, which appear as early
landscapes in the Mediterranean area and elsewhere (Bolivia, Peru etc.) as 5000 BP (Wilkinson, 1999), are threatened in many areas due to
have been abandoned, often for social-economic reasons (Koulouri and deterioration and abandonment (Pietsch and Mabit, 2012; Varisco,
Giourga, 2007, Tarolli et al., 2014). Cultivation abandonment can have 1991; Vogel, 1988). Overall, in most other cases adverse environmental
⁎
Corresponding author.
E-mail address: moraetis@yahoo.gr (D. Moraetis).
https://doi.org/10.1016/j.geoderma.2019.114152
Received 26 March 2019; Received in revised form 15 December 2019; Accepted 18 December 2019
0016-7061/ © 2019 Elsevier B.V. All rights reserved.
D. Moraetis, et al. Geoderma 363 (2020) 114152
impacts have been shown to follow terrace abandonment, including 41.7 °C (Celsius), respectively and at the Saiq station from 3.6 to
examples in southern Peru (Londoño, 2008), on the Greek islands 31.2 °C. The average daily fluctuations for December 2018 and July
(Koulouri and Giourga, 2007; Moraetis et al., 2015) and in Spain 2018 at Seeb range from 17 to 27 °C and 32 to 39 °C, respectively. The
(Lasanta et al., 2001). data for the same months at Saiq show a range from 7 to 18 °C and from
In Oman and on the southern flanks of the Jabal Akhdar, agriculture 22 to 31 °C. The evapotranspiration for 9 different crops and palm trees
and eco-tourism in mountainous high elevation areas (up to 1980 m) at a typical oasis in the northern Jabal Akhdar (Balad Seet) showed a
offer significant economic opportunities. However, some land aban- reference evapotranspiration of 2393 mm year−1 (Siebert et al., 2007).
donment is still taking place, which threatens agricultural sustainability The final exhumation of the Jabal Akhdar Dome occurred during the
and cultural landscapes (Al-Rawahi et al., 2014, Buerkert et al., 2010). Eocene to Miocene (Hansman et al., 2017; von Grobe et al., 2019).
In 2008, Luedeling and Buerkert identified the water shortages in the Neoproterozoic sedimentary rocks are separated by an angular un-
mountain oases of Oman as a contributing factor to soil terrace aban- conformity with Permo-Mesozoic rocks of the Hajar Supergroup. The
donment. The same study also indicates that shifts toward irrigation former consist of mostly sandstone, siltstone, greywacke and limestone,
demanding crops (roses and fruits) and precipitation pattern changes while the latter mostly consists of carbonates (Fig. 2c, d; Béchennec
have contributed to land abandonment. In an extremely arid environ- et al., 1992). The two villages are located immediately below the an-
ment like Oman, terrace abandonment leads to soil desertification, gular unconformity (Fig. 2c, d). Groundwater penetrates the limestone
erosion and terrace wall collapse (Luedeling, 2007). of the Hajar Supergroup and forms springs along the contact with im-
With time, such techniques of soil agriculture in extreme environ- permeable underlying siliciclastic Neoproterozoic rocks. These springs
ments are in danger of falling into oblivion especially when these areas form the basis for several permanent settlements within the Jabal
are getting abandoned by young generation (Faccini et al., 2018). In- Akhdar Dome, including the two studied sites.
vestigations of ancient soil terrace landscapes contribute to knowledge Hadash Village (Fig. 2a) is situated in the upper reaches of Wadi
of food production among ancient civilizations (Beach et al., 2002; Mistal which drains to the NW, and Wijma Village (Fig. 2b) is located in
Bevan and Conolly, 2011; Bevan et al., 2013: Charbonnier, 2017; the upper reaches of the Wadi Sahtan which drains to the NE. The
Harrower and Nathan, 2018). Understanding ancient agricultural ter- Hadash soil terraces have been built on the top of partly tawny dolomite
racing requires an understanding of how soils form, including their (not shown in Fig. 2c) and pinkish siltstone of the Mistal Formation
regolith provenance. Natural soil development in countries like Oman is while the terraces of Wijma rest on the top of pinkish siltstone of the
often poor due to aridity, resulting in a lack of vegetation and low TOC Mistal Formation (Fig. 2d).
values, while extreme precipitation events often result in soil erosion
(Ministry of Regional Municipalities, Environment, and Water 3. Methods
Resources, 2005). Nagieb et al. (2004) suggested that in Oman moun-
tainous areas, the sediment/regolith was brought in to soil terraces 3.1. Sampling strategy
from nearby wadis after heavy rainfalls. However, no studies have
demonstrated the regolith provenance in mountainous soil terraces in Three soil pits were excavated in different abandoned soil terraces
Oman. Details of such agricultural measures for terrace construction in in Hadash and three in Wijma. We characterized and sampled each soil
harsh environments are significant for re-evaluation of ancient agri- profile horizon. Hadash soil samples are labelled “H” and Wilma soil
cultural practices. The regolith provenance in such anthropogenically samples “W”. In addition, we sampled the bedrock of the two villages,
influenced landscapes is not only of local interest, but also of global tawny dolomite from Hadash and pinkish siltstone from Wijma.
importance. Such features have a special significance to settlements in Furthermore, we collected nine samples from different localities of
extreme environments. Wadi Mistal and Wadi Sahtan, representing unconsolidated, fine-
Since the lack of pedogenesis due to high aridity, mountainous grained sediments (possible soil parent rock regolith) and soils. Three
terraced soils in Oman constitute a significant sedimentary record unconsolidated sediments (regolith) included one lacustrine sediment
under an imminent risk of loss. Moreover, terrace abandonment leads to from Wadi Mistal, one lacustrine sediment from Wadi Sahtan, and one
adverse impacts on the soil quality especially on those marginal areas. recent dam deposit (dam construction occurred between 2003 and 4
Herein, we demonstrate the unique natural value of those areas by based on past Google Earth imagery) from Wadi Mistal. The soils were
studying the soil provenance and the complex pedogenesis in two collected from nearby agricultural soil terraces, two abandoned (Hubra
mountain oases of Oman. We also add new findings on the impact of and Al Sab soils) and two cultivated (soil terraces in Wadi Abyad and
terrace abandonment on soil quality. To achieve the previous, we soil in Wadi Sahtan) for comparison with the soil of the abandoned
present geochemical, physical, mineralogical, microscope and 14C terraces of Hadash and Wijma. Moreover, we identified an abandoned
dating evidence of the regolith provenance and soil conditions. old Wijma settlement ~100 m above present-day Wijma. We collected
in-situ developed soil and parent rock from soil terraces of this aban-
2. Physical setting doned locality. Google Earth was used to spot the existence of farms
nearby our study areas where we could collect soil. The samples were
The villages of Hadash and Wijma are located within the Jabal examined for geochemical and mineralogical comparison with the ter-
Akhdar Dome of the Central Oman Mountains (Fig. 1). The climatic raced soils from Hadash and Wijma. One particular soil sample from
conditions are characterized by general aridity throughout the year Hubra Village was also collected because it represents a well-developed
with some scattered and erratic rainfalls and storms (Abdalla and Al- calcrete in a paleosol. This sample is considered significant for com-
Abri, 2011; Shahalam, 2001). For the assessment of the climatic con- parison with our study sites with respect to our 14C results (see Section
ditions we have used data from meteorological stations of the Global 3.3). Hadash and Hubra soils are situated in the same drainage basin.
Historical Climatology Network (GHCN) of the National Oceanic and The Hubra soil occurs downstream from Hadash along Wadi Mistal. In
Atmospheric Administration (NOAA). Full details of the meteorological total 35 soil and rock samples were collected to compare geochemistry
stations are provided in the methodology section. The average recorded and mineralogy (coordinates and elevation are provided in Table 2).
rainfall between 1991 and 2018 at the Seeb station (elevation: 15 m)
was 137 mm, while at the Saiq station (elevation: 1755 m) it amounted 3.2. Geochemical and physical sample characterization
to 175 mm. Most of the precipitation events occurred during winter
(December to March) except for some cyclons (Kotwicki and Al The samples from Hadash and Wijma were analyzed for their pH
Sulaimani, 2009). The minimum and maximum temperatures recorded values and electrical conductivity in the supernatant solution of a
at the Seeb station for the period 1991–2018, range from 15.2 to mixture 1:2 of soil to water. Sand silt and clay were measured following
2
Fig. 1. Regional Geology outline of the Northern Oman, meteorological stations (Saiq, Semail, Seeb). The studied sites Hadash and Wijma are also shown. Modified
after Béchennec et al. (1993), Moraetis et al. (2018) and Mattern et al. (2018).
the method of Bouyoucos (1962). The TOC content was determined by INTCAL 13 was used (Reimer et al., 2013). The stable isotopes 13C and
18
acid dichromate digestion technique (Walkley and Black, 1934). Ag- O were also measured and reported along with 14C by BETA Analytic
gregate stability of the soil samples was measured using the method of Inc. The 13C and 18O were presented relative to Vienna Pee-Dee Be-
Stamati et al. (2013). Aggregates were collected through wet sieving lemnite standard (VPDB).
with 2 mm, 1 mm, 0.250 mm and 0.063 mm mesh. The aggregate
concentration was corrected for the sand content. The weight of each 3.4. Spring water analysis and historical precipitation data
fraction was measured after drying in an oven at 38 °C.
Niton XL3t Goldd (Thermo Fischer) X-Ray fluorescence analyses Water samples were collected from two springs in Hadash and
were carried out with samples from Hadash and Wijma along with the Wijma on April 29th, 2016 and on June 6th, 2016, respectively.
unconsolidated (regolith) and soils samples collected from Wadi Mistal, Samples were filtered through 0.45 μm pore size Whatman paper and
Wadi Sahtan and old Wijma Village. X-ray diffraction (XRD) analyses several physicochemical parameters (pH, electrical conductivity, dis-
were performed for mineralogical analysis with the X-Pert Pro instru- solved oxygen, oxidation-reduction potential) were measured on site
ment (Panalytical). using Aquaread AP800. Bulk chemical analyses were carried out in the
filtered samples with ICP-OES (Varian 710-ES from Varian Inc.) at
3.3. Radiocarbon dating Sultan Qaboos University in the Central Analytical and Applied
Research Unit (CAARU).
Three calcrete samples were dated from soil horizons in Hadash, Precipitation data were downloaded from the National Oceanic and
Wijma and Hubra. The sample of Wijma was from pit W3 collected at a Atmospheric Administration (NOAA). The three selected meteor-
depth between 9 and 21 cm and for Hadash was from pit H3 and from a ological stations are Saiq (elevation 1755 m, 23.07°N, 57.65°E), Seeb
depth between 38 and 56 cm. The sample from Hubra was collected (elevation 15 m, 23.59°N, 58.28°E) and Semail (elevation 127 m,
from a depth of 5 cm. The respective soil profiles and samples are 23.32°N, 57.95°E). These stations were selected because they are lo-
shown in Fig. 3a–f. The samples were dated with an Accelerator Mass cated at very different elevations that span the study area and because
Spectrometer (AMS) by BETA Analytic Inc. For all the samples the bulk of their complete data sets. The Precipitation Index (PI) was calculated
carbonates were measured by acid etching. The Libby half-life was used using the equation (x − xaverage)/xaverage, where x is the yearly pre-
for the calculation of the raw 14C yr BP. The calibrated years of the raw cipitation and xaverage the average rainfall in the period 1990–2018, for
14
C data were derived from BetaCa l3.21, and for calibration the curve the Saiq and Seeb stations and in the period 2003–2018 for the Semail
3
Fig. 2. Geology outline of the studied areas, (a) Hadash, (b) Wijma, (c) Hadash geological map, (d) Wijma geological map. Map after Beurrier et al. (1986b) and Rabu
et al. (1986).
station. (Varian 710-ES from Varian Inc.) and Aquaread AP800.

The principal component analysis was applied for the bulk chemical
analysis (oxides) for all the 35 samples collected in the present study.
3.5. Statistical analysis The number of the principal components were determined by the se-
lecting of the eigenvalue with score > 1. The first two principal com-
We have performed analysis in triplicates for the TOC measure- ponents corresponded to this condition. A correlation matrix was con-
ments and for the soil aggregate analysis, for every soil horizon and structed prior to the PCA.
standard deviation and mean values were calculated. Standard devia-
tion of the XRF-results is supplied automatically by the Niton XL3t
Goldd, while for the soil EC is determined by the instrumentation 4. Results
software (Aquaread AP800). The mean and the standard deviations
were fed into PAST 3 software for the T-test between different soil 4.1. Hadash and Wijma soils and their physical characteristics
horizons for CaO and SiO2 content, EC values and aggregates content
with a null hypothesis significance of 95%. The water data analysis is According to FAO (2006), the soils from both sampling areas re-
presented with a standard deviation supplied by the ICP-OES software present anthrosols. Identified soil horizons are shown in Fig. 4a, b and
4
Table 1
Soil horizon characterization, soil pH, soil electrical conductivity (EC), soil texture, Munsell color (dry) and organic C content from Hadash and Wijma. The
measurement accuracy is 1% for EC and less than 5% for organic content.
Depth (m) Horizon characterization pH EC (μS/cm) Texture Munsell color (dry) Organic C content (g kg−1)
H1 (0–5) AC 7.85 46 loam 2.5YR 5/3 4.90

H1 (5–20) C1 7.78 124 loam 2.5YR 5/3 5.36
H1 (20–38) Cm2 7.89 69 sandy loam 2.5YR 5/3 2.33
H1 (38–56) Cm3 7.83 59 sandy clay loam 2.5YR 6/4 2.33
H2 (0–5) AC 8.11 42 loam 10YR 6/2 9.79
H2 (5–20) C1 7.88 48 clay loam 2.5YR 5/2 5.83
H2 (20–38) Cm2 7.48 319 sandy loam 2.5YR 5/3 4.90
H2 (38–46) Cm3 7.55 601 clay loam 2.5YR 6/2 5.13
H2 (46–65) Cm4 7.62 1680 sandy loam 2.5YR 5/3 5.60
H3 (0–5) AC 7.88 85 loam 2.5YR 5/3 11.43
H3 (0–16) C1 7.65 84 sandy loam 2.5YR 5/5 9.56
H3 (16–33) C2 7.89 167 loam 10YR 5/3 7.23
W1 (0–9) AC 7.91 189 silty loam 10YR 3/1 4.65
W1 (9–24) C1 7.37 77 sandy clay loam 2.5YR 4/1 4.23
W1 (24–55) C2 7.25 490 clay loam 2.5YR 5/2 3.81
W1 (55–70) C3 7.50 859 loam 2.5YR 5/2 3.38
W2 (0–9) AC 7.95 81 loam 2.5YR 4/2 10.36
W2 (9–30) Cm1 7.95 66 sandy loam 2.5YR 4/1 8.24
W2 (30–60) C2 7.61 426 loam 2.5YR 5/3 4.65
W2 (60–73) C3 7.95 206 clay loam 2.5YR 6/3 2.11
W3 (0–9) AC 8.02 107 sandy loam 2.5YR 4/1 5.29
W3 (9–21) Cm1 7.79 134 sandy clay loam 2.5YR 4/3 5.07
W3 (21–39) Ck2 7.68 143 loam 2.5YR 5/2 4.65
W3 (39–47) Cr3 7.95 85 clay loam 2.5YR 5/3 3.81
pH, electrical conductivity, texture and TOC content are shown in and color is mainly reddish-brown (2.5YR 5/3) with some variability in
Table 1. The pH for the Hadash soils range between 7.55 and 8.11, and different horizons (Table 1).
the electrical conductivity (EC) is mostly < 300 μS cm−1, except for the The pH of Wijma soils varies between 7.3 and 8, and the EC is
Cm3 and Cm4 horizons in profile H2 where it ranges between 601 and mostly < 300 μS cm−1, while some higher values appear in deeper
1680 μS cm−1. The texture of the Hadash soils varies in different soil horizons of W1 and W2 soil pits (490–859 and 426 μS cm−1, respec-
horizons (Table 1) however, the average clay silt and sand contents is of tively). The texture is identical to that of the Hadash soils with average
22%, 31%, 47%, respectively, denoting to loam texture. The TOC clay, silt and sand of 24%, 31% and 45%, respectively (loam). The dry
content is less than 6 g kg−1, except for the top horizon in H2 Munsell color is dark reddish to weak red (2.5YR 4/1, 4/2) with some
(9.8 g kg−1) and the shallow profile H3 (11 g kg−1). The dry Munsell variability as it is in Hadash soils (Table 1). The T-test was applied to
Table 2
Water Stable Aggregates results (WSA %) and Mean Weight Diameter (mm). The weight distribution in the different sieves is shown. Parentheses in the first column
shows standard deviation of the measured weight. Letters denoting the T-test results. Results with the same lowercase letter within the same row are not significantly
different (p < 0.05). Results of the same uppercase letter within the same column is not significantly different at p ≤ 0.05.
Hadash Wijma
Fractions (standard deviation) Η1 Η2 Η3 W1 W2 W3
Depth/horizon (0–5 cm)/AC (0–5 cm)/AC (0–5 cm)/AC (0–5 cm)/AC (0–9 cm)/AC (0–9 cm)/AC
> 2000 μm (1.7) 12.8 a 10.8 a 7.1b A 13.6 aA 6.2 b 7.2 b
2000–1000 μm (1.1) 32.3 a 5.8 b 4.6 b B 21.4 cB 4.9 b 3.8 b
1000–250 μm (1.7) 18.0 a 25.1 b 8.1 c C 18.5 aD 16.1 a 9.9 c
250–63 μm (2.2) 15.1 a 11.6 b 15.2 a D 13.2 ab E 12.9 ab 9.8 b
< 63 μm (1.0) 6.5 a 7.0 a 16.0 b E 14.8 bF 13.8 b 12.0 b
MDW (mm) (0.03) 0.46 a 0.45 a 0.28 a F 0.48 aH 0.30 a 0.27 a
WSA% (> 250 μm) (0.8) 63.1 a 41.8 b 19.8 c G 53.5 dJ 27.1 e 21.0 f
Depth/horizon (5–16 cm)/C1 (5–9 cm)/AC

> 2000 μm (2.4) 7.3 a A 12.2 a A
2000–1000 μm (0.7) 3.4 a B 6.0 b C
1000–250 μm (3.8) 10.0 a C 18.7 b D
250–63 μm (5.6) 11.2 a D 13.7 a E
< 63 μm (1.1) 14.1 a E 20.8 b G
MDW (mm) (0.03) 0.27 a F 0.48 a H
WSA% (> 250 μm) (3.8) 20.7 a G 36.9 b K
Depth/horizon (16–33 cm)/C2 (9–19 cm)/C1

> 2000 μm (2.3) 9.3 a A 11.3 a A
2000–1000 μm (1.2) 5.3 a B 6.8 a C
1000–250 μm (1.0) 17.6 a H 18.7 a D
250–63 μm (2.5) 11.7 a D 5.2 b L
< 63 μm (1.2) 11.0 a J 7.3 b M
MDW (mm) (0.04) 0.38 a F 0.43 a H
WSA% (> 250 μm) (3.4) 32.1 a K 36.8 a K
5
Hadash Wijma Hubra
(a) (b) (c)
(d) (e) (f)
Fig. 3. (a) Soil pit in Hadash (profile H1), (b) Soil pit in Wijma (profile W3), (c) Hubra calcrete, (d) Hadash soil sent for 14C, nodular aggregation (e) Wijma soil for
14
C analysis. The calcrete horizon was poorly consolidated but still small concretions are present, (f) Hubra soil for 14C with elongated nodules.
H2 H1
H3 AC 5 AC
Slope 9% 5 C1
AC 20 C1 20
5 Cm2 Cm2
C1 38 38
13 Cm3
C2 Cm3 56
33 46
a) Hadash Cm4
R-Tawny 65
R-Tawny
Dolomite
Dolomite
W1
AC
9
W2 C1
24
AC C2
W3 9 55
Slope 49% Cm1 C3
AC 30 77
9
Cm1 C2
21 60
b) Wijma Ck2 C3
39 73
Cr3
47
R-Mistal
Pinkish
Siltstone
Fig. 4. a) Characterization of Hadash soil profiles, b) Characterization of Wijma soil profiles. For both areas the cliff slopes are shown.
6
Table 3
Major elements analysis with XRF from Hadash and Wijma soils and bedrock, possible parent rocks, other soil terraces and in situ soil from the old Wijma soil. Loss of Ignition (LOI) is also shown in the last column.
Parentheses present standard deviation.
D. Moraetis, et al.
Elevation (m) Coordinates Codes for soil samples and description for MgO% Al2O3% SiO2% P2O5% SO2% K2O% CaO% TiO2% Fe2O3% LOI%
bedrock, possible sources and other soils
Soil-Hadash (wadi Mistal) 1,469 23.128672°N, H1 (0–5 m) 2.0 (0.3) 7.3 (0.1) 23.1 (0.1) 0.9 (0.02) 0.4 (0.01) 2.4 (0.03) 33.9 (0.1) 0.3 (0.01) 6.4 (0.02) 20.0
57.755753°E H1 (5–21 m) 1.2 (0.4) 6.0 (0.1) 18.3 (0.1) 0.7 (0.02) 0.3 (0.01) 1.9 (0.03) 27.1 (0.1) 0.3 (0.01) 5.4 (0.02) 43.4
H1 (20–38 m) 1.0 (0.3) 6.7 (0.1) 20.7 (0.1) 1.0 (0.02) 0.4 (0.01) 2.1 (0.03) 38.0 (0.2) 0.3 (0.01) 5.7 (0.02) 20.5
H1 (38–57 m) 1.1 (0.4) 6.0 (0.1) 17.9 (0.1) 0.8 (0.02) 0.4 (0.01) 2.0 (0.03) 42.6 (0.2) 0.3 (0.01) 5.9 (0.02) 18.9
1,469 23.128535°N, H2 (0–5 m) 1.7 (0.3) 8.0 (0.1) 25.1 (0.1) 1.0 (0.02) 0.4 (0.01) 2.7 (0.03) 30.6 (0.1) 0.4 (0.01) 7.5 (0.02) 20.0
57.755573°E H2 (5–20 m) 1.3 (0.4) 7.9 (0.1) 24.0 (0.1) 0.8 (0.02) 0.4 (0.01) 2.6 (0.03) 29.7 (0.1) 0.3 (0.01) 7.5 (0.02) 23.0
H2 (20–38 m) 1.2 (0.3) 7.6 (0.1) 23.7 (0.1) 0.8 (0.02) 0.4 (0.01) 2.5 (0.03) 31.5 (0.1) 0.4 (0.01) 6.9 (0.03) 22.8
H2 (38–46 m) – 8.8 (0.1) 22.8 (0.1) 0.7 (0.02) 0.4 (0.02) 3.1 (0.04) 33.1 (0.2) 0.4 (0.01) 8.7 (0.03) 20.5
H2 (46–65 m) 1.4 (0.4) 6.9 (0.1) 19.0 (0.1) 0.6 (0.02) 0.4 (0.01) 2.6 (0.03) 36.8 (0.1) 0.3 (0.01) 5.8 (0.02) 21.1
1,466 23.129288°N, H3 (0–5 m) 1.6 (0.3) 7.4 (0.1) 25.2 (0.1) 0.8 (0.02) 0.3 (0.01) 2.6 (0.03) 29.8 (0.1) 0.3 (0.01) 7.4 (0.02) 20.0
57.754285°E H3 (5–16 m) 1.4 (0.3) 7.6 (0.1) 25.5 (0.1) 0.7 (0.02) 0.3 (0.01) 2.5 (0.03) 30.1 (0.1) 0.4 (0.01) 7.0 (0.02) 21.5
H3 (16–33 m) 1.9 (0.3) 7.5 (0.1) 25.1 (0.1) 0.7 (0.02) 0.3 (0.01) 2.5 (0.03) 29.1 (0.1) 0.3 (0.01) 6.9 (0.02) 21.4
Soil-Wijma (wadi Sahtan) 1,238 23.220374°°N, W1 (0–9 m) 1.4 (0.3) 10.7 (0.1) 31.8 (0.1) 1.5 (0.02) 0.5 (0.01) 3.5 (0.06) 22.0 (0.2) 0.3 (0.01) 8.0 (0.02) 20.0
57.307519°E W1 (9–24 m) – 9.4 (0.1) 30.2 (0.1) 1.1 (0.02) 0.8 (0.01) 3.3 (0.05) 23.5 (0.2) 0.4 (0.01) 9.7 (0.03) 19.2
W1 (24–55 m) 1.6 (0.3) 10.0 (0.1) 29.4 (0.1) 1.0 (0.02) 0.7 (0.01) 3.0 (0.05) 23.5 (0.1) 0.4 (0.01) 8.1 (0.02) 19.8
W1 (50–70 m) 1.3 (0.4) 9.8 (0.1) 27.9 (0.1) 0.7 (0.02) 0.7 (0.01) 2.9 (0.05) 26.3 (0.2) 0.3 (0.01) 7.6 (0.02) 20.2
1,230 23.220316°N, W2 (0–9 m) 1.0 (0.4) 8.9 (0.2) 27.7 (0.1) 1.4 (0.02) 0.5 (0.02) 2.7 (0.06) 29.9 (0.2) 0.2 (0.01) 8.2 (0.03) 20.0
57.307553°E W2 (9–30 m) – 7.4 (0.1) 26.7 (0.1) 1.3 (0.02) 0.5 (0.01) 3.1 (0.04) 28.6 (0.2) – 6.9 (0.02) 21.9
W2 (30–60 m) 1.1 (0.3) 7.1 (0.1) 26.9 (0.1) 0.8 (0.02) 0.4 (0.01) 2.4 (0.04) 28.8 (0.1) 0.3 (0.01) 6.9 (0.02) 21.3
W2 (60–73 m) 1.0 (0.3) 9.0 (0.1) 27.8 (0.1) 0.7 (0.02) 0.4 (0.01) 2.9 (0.05) 28.3 (0.2) 0.4 (0.01) 7.3 (0.02) 17.3
1,227 23.220217°N, W3 (0–9 m) 1.8 (0.3) 8.9 (0.1) 26.7 (0.1) 1.2 (0.02) 0.4 (0.01) 2.8 (0.05) 28.6 (0.1) 0.3 (0.01) 7.1 (0.02) 20.0
7
57.307638°E W3 (9–21 m) 1.1 (0.3) 9.2 (0.1) 28.4 (0.1) 0.9 (0.02) 0.4 (0.01) 3.0 (0.05) 28.5 (0.1) 0.3 (0.01) 8.6 (0.02) 16.7
W3 (21–39 m) – 9.5 (0.1) 28.3 (0.1) 0.8 (0.02) 0.5 (0.01) 3.0 (0.05) 27.9 (0.1) – 8.2 (0.02) 19.5
W3 (39–47 m) – 11.5 (0.1) 30.3 (0.1) 0.7 (0.02) 0.5 (0.01) 3.5 (0.04) 29.5 (0.2) 0.4 (0.01) 9.6 (0.03) 11.9
Bedrock
Bedrock-Hadash 1,466 23.128898°N, Tawny dolomite 3.0 (0.4) 1.1 (0.1) 9.3 (0.1) – 1.1(0.01) 0.7 (0.04) 30.2 (0.2) 0.1 (0.01) 3.3 (0.02) 42.11
57.755165°E
Bedrock-Wijma 1,227 23.220692°N, Mistal Fm. Pinkish siltstone – 12.4 (0.1) 72.0 (0.1) 0.1 (0.02) 0.1(0.01) 4.8 (0.04) 0.2 (0.02) 1.3 (0.01) 1.1 (0.02) 5.13
57.307213°E
Possible soil parent rock
Lacustrine sediments in 523 23.301710°N, Late Pleistocene sediments – 3.7 (0.1) 32.5 (0.1) – 0.1 (0.01) 1.2 (0.04) 29.0 (0.2) 0.4 (0.01) 3.7 (0.03) 21.1
wadi Mistal 57.693995°E
Dam sediments in wadi 581 23.262264°N, Recently deposited sediments – 8.5 (0.1) 47.6 (0.1) 0.2 (0.02) 0.4 (0.01) 2.8 (0.05) 16.5 (0.2) 0.7 (0.01) 7.5 (0.02) 11.11
Mistal 57.706101°E
Lacustrine sediments in 802 23.241751°N, Possibly late Pleistocene sediments – 2.5 (0.1) 14.8 (0.1) – 0.1(0.01) 1.0 (0.04) 51.3 (0.2) 0.1 (0.01) 1.8 (0.02) 25.1
wadi Sahtan 57.335509°E
Other soil terraces
Soil-Hubra 179 23.495974°N, Abandoned soil terraces highly cemented with 1.3 (0.4) 3.3 (0.1) 30.4 (0.1) 0.4 (0.02) 0.2 (0.01) 1.1 (0.03) 30.9 (0.3) 0.4 (0.01) 3.7 (0.02) 22.3
57.837003°E calcite
Soil-wadi Abyad 210 23.461260°N, Cultivated soil terraces 2.9 (0.3) 5.1 (0.1) 36.1 (0.1) 1.2 (0.02) 0.2 (0.02) 1.1 (0.03) 22.8 (0.1) 0.4 (0.01) 5.5 (0.02) 19.3
57.666289°E
Soil-As Sab 1,693 23.218183°N, Abandoned soil terraces 0.9 (0.2) 3.7 (0.1) 24.5 (0.1) 0.3 (0.02) 0.3 (0.01) 1.3 (0.04) 33.1 (0.2) – 3.2 (0.02) 28.6
57.208570°E
(continued on next page)
Geoderma 363 (2020) 114152
LOI% the EC, and the result (p = 0.71) shows that the mean EC for the Ha-
23.8
14.3
15.1
dash (298 μS cm−1) and Wijma soils (238 μS cm−1) are similar. The T-
5.6 (0.02)
test for the mean TOC content for the Hadash (6.2 g kg−1) and Wijma
8.3 (0.02)
8.0 (0.03)
soils (5.0 g kg−1) showed that they are also statistically similar
Fe2O3%
(p = 0.3).
AC and C horizons (Fig. 4a, b) are mainly present at both sites. In
several cases, calcite cements of the soil horizons were observed. The
0.5 (0.01)
0.6 (0.01)
0.7 (0.01)
TiO2%
Hadash soil profiles exhibit extensively cemented blocky soil as shown

in Fig. 3a and d of the Hadash profile H2. Cementation was pronounced
in the C horizon for the Hadash soils (Fig. 4a). Cement was also ob-
17.9 (0.3)
14.4 (0.2)
9.0 (0.2)
served in Wijma soils, but there the soil is more friable and the blocky
CaO%
texture was unstable (Fig. 3b, e). The Hadash soils appear slightly
shallower in depth (33–65 cm) than those of Wijma (47–77 cm), while
2.3 (0.03)
3.0 (0.05)
3.8 (0.04)
terraces of the latter have been constructed at a much steeper slope

K2O%
(~49° vs. ~9°) (Fig. 4). In the H2 and H3 profiles of Hadash, the
bedrock (tawny dolomite) was encountered in the soil pit excavation at
a depth of 33–60 cm, respectively. The pinkish siltstone of the Mistal
0.4 (0.01)
0.2 (0.01)
0.2 (0.01)
Formation was met in the W3 profile of Wijma at a depth of 47 cm.

SO2%
Wet aggregate analyses are shown in Table 2. The T-test was applied
and the results are shown in the same table. The water stable aggregate
0.7 (0.02)
0.3 (0.02)
0.2 (0.02)
percentage (WSA) is < 42% in 8 out of the 10 the soil horizons of

P2O5%
Hadash and Wijma.

35.5 (0.1)
44.0 (0.1)
48.9 (0.1)
4.2. Results on bulk chemical analysis and dating

SiO2%
The CaO content varies for both the Hadash soils and the Wijma
soils (Table 3). The average values are 32.7% ( ± 4.5%) and 27.1%
10.9 (0.1)
6.6 (0.1)
7.8 (0.1)
Al2O3%
( ± 2.7%) for Hadash and Wijma, respectively. The T-test showed that
they are not statistically different. The average SiO2 average con-
centrations in Hadash and Wijma soils are 22.5% ( ± 2.8%) and 28.5%
1.3 (0.4)
1.1 (0.3)
( ± 1.6%), respectively, which are also not statistically different

MgO%
(P < 0.05). The bedrock in the Hadash area is dolomite with an

–
identical CaO abundance in the Hadash soil. However, the MgO

(3.3 ± 0.4%) and Al2O3 (1.1 ± 0.1%) contents differ from that of the
Mu′aydin Fm, siltstone and carbonate beds
overlying soils (3.3 ± 0.4% for Mg). The pinkish siltstone bedrock in
Codes for soil samples and description for
bedrock, possible sources and other soils
Wijma (Mistal Fm.) displays a very low level of CaO (0.2 ± 0.02%),
and a high content of SiO2 (72 ± 0.1%).
The three sediments (regolith) considered as possible parent rock,
the lacustrine sediments from Wadi Mistal and Wadi Sahtan and the
recently deposited dam sediments, show a high variability in the SiO2
Abandoned soil terraces
Cultivated soil terraces
and CaO concentration (Table 3). The lacustrine sediments from Wadi
Mistal exhibit SiO2 and CaO concentrations, close to those observed in
the Hadash soil terraces. The recent dam sediments from Wadi Mistal,
which occur in a distance of 5.2 km from the lacustrine sediments of
Wadi Mistal, show a low CaO content and a high SiO2 content com-
pared to the Hadash soil. The lacustrine sediments from Wadi Sahtan
show a significant difference in CaO and SiO2 concentration, with CaO
concentrations to be much higher than those observed in the Wijma
soils (Table 3).
23.311118°N,
23.224084°N,
23.224084°N,
The other cultivated soils from Wadi Sahtan and Wadi Abyad and
57.323272°E
57.308698°E
57.308698°E
Coordinates
the uncultivated soils from Hubra and the terraces of Al Sab show
concentrations of SiO2 and CaO which are similar to those of the soils
from Hadash and Wijma. The presently cultivated soils from Wadi
Sahtan represent an exception with lower CaO concentration compared
Elevation (m)
to those from Hadash and Wijma (Table 3). Generally, the CaO content
is low in the cultivated soil terraces. The uncultivated in-situ soil from
1,312
1,312
597
the abandoned old Wijma Village and the parent rock display similar
content of SiO2 (44% for soil and 48 ± 0.1% for regolith). The CaO
Soil in the old Wijam village
concentration in the in-situ soil is slightly higher than that of the un-
Example of insitu soil and
Bedrock in the old Wijam
derlying parent rock (Table 3).

Table 3 (continued)
Principal Component Analysis (PCA) of the 35 rocks and soils shows

Soil-wadi Sahtan
that the first three Principal Components (PC) are capturing 80% of the
parent rock
variability. CaO and SiO2 concentrations display the highest variability

village
as expected. The scatter plot shows that soils from Hadash and Wijma
are grouped together regarding PC1 with scores −1.5–1.5 (numbers
1–24 in Fig. 5a). The PC1 loading plot exhibits the SiO2 concentration
8
b)
a)
Component 1
c)
1-12: Hadash samples
13-24 :Wijma samples
25-26: Hadash and Wijma bedrock
27-29: Hadash, Wijma lacustrine
sediments and dam sediments
30-33: Other soil terraces
34-35: Soil and bedrock from the
old Wijma village
Component 2
Fig. 5. PCA analysis of major elements a) Score scatter plot, b) PC1 loadings, c) PC2 loadings.
and CaO negative correlation which is the most important factor ex- they are absent from the other soil samples. Illite is common in the soils
plaining 45% of the data variability (Fig. 5b). The PC2 loading plot of Hadash, Wijma and Hubra as well as in the dam sediments of Wadi
explains 24% of the variability and separates Hadash and Wijma soils Mistal. The weathering-resistant mineral of quartz and the less resistant
from the other samples due to higher P2O5 and Fe2O3 concentrations albite are present in most of the rocks and soils. Notable is the presence
(Fig. 5c). PC3 captures 10% of the variability, and that is explained of lizardite in the soils of Hadash, Wijma, Hubra and in the soil terraces
mainly due to higher MgO concentrations in a few samples like the in Wadi Abyad. Muscovite and chlorite have been identified in most of
bedrock in Hadash. the soils and rocks except for the soils in Hadash and Wijma.
The mineralogy shown in Table 4 documents the existence of ver- The carbon-14 dating results are shown in Table 5. Hadash and
miculite and montmorillonite in Hadash and Wijma soils only, while Hubra soils reveal a similar age of parent material with
Table 4
Mineralogy results from Hadash and Wijma soils and bedrock, possible parent rocks, other soil terraces and in situ soil from the old Wijma soil.
Qu Al K-f An Ca Do Mu Cl Chl Il Ve Mo Ka Li Ch Pa
Soil-Hadash (wadi Mistal) ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Soil-Wijma (wadi Sahtan) ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Bedrock
Bedrock-Hadash ✓ ✓ ✓ ✓ ✓ ✓ ✓
Bedrock-Wijma ✓ ✓ ✓ ✓ ✓
Possible parent rocks

Lake sediments in wadi Mistal ✓ ✓ ✓ ✓ ✓ ✓ ✓
Dam sediments in wadi Mistal ✓ ✓ ✓ ✓ ✓ ✓ ✓
Lake sediments in wadi Sahtan ✓ ✓ ✓ ✓ ✓
Other soil terraces

Soil-Hubra ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Soil terraces in wadi Abyad ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓
Soil-As Sab ✓ ✓ ✓ ✓ ✓
Soil in wadi Sahtan ✓ ✓ ✓ ✓ ✓ ✓
Example of insitu soil and parent rock

Soil in the old Wijam village ✓ ✓ ✓ ✓ ✓ ✓
Bedrock in the old Wijam village ✓ ✓ ✓ ✓ ✓ ✓
Qu: quartz, Al: albite, K-f: K-feldspar, An: anorthite, Ca: calcite, Do: dolomite, Mu: muscovite, Cl: clinochloro, Chl: chlorite Il: illite, Ve: vermiculite, Mo:
Montmorillonite Li: lizardite, Ch: chrysotile, Pa: palygorskite.
9
Table 5
14
Age results from C analysis in calcrete samples from Hadash, Wilma and Hubra. Stable isotopes are shown. 2σ calibration results are also shown.
Sample code Laboratory code Calibration 2σ BC Calibrated age ka BP δ13C o/oo δ18O o/oo
H1 38–57 cm Beta−452,901 12060–11815 13887 ± 40 BP −3.9 −2.2

W3 9–21 cm Beta−452,902 8280–8225 10193 ± 30 BP −5.6 −3.5
Hubra Beta−462,839 11324–11151 13186 ± 30 BP −0.7 −3.9
(a) (b) (c) (d)
1000 µm 100 µm 200 µm 200 µm
(e) (f) (g) (h)
200 µm 200 µm 200 µm

1000 µm
(i) (j) (k) (l)
1000 µm
100 µm 200 µm 200 µm
Fig. 6. (a–d) Hadash thin sections in PPL (a, b) and XPL (c, d), in the presentation order, microsparite with wrinkled cracks and voids with extraclasts of possibly
slate, Microcodium with thin thin rim, ostracod fragment with wavy ornament, ostracod with radial sweeping extinction, (e–h) Wijma thin section in PPL (e, f) and
XPL (g, h), in the presentation order we see layered micritic intraclast, Microcodium with anhedral calcite, ostracod fragment, large ostracod fragment with radial
sweeping extinction, (i–l) Hubra thin sections in PPL (i) and XPL (j–l), angular to rounded clastic material in a displacive texture with clotted micrite, subangular
quartz and angular feldspar, bioplastic fragment of possible ostracod (k), and red algae (l). (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
13.887 ± 40 yr cal BP and 13.186 ± 30 yr cal BP, respectively. The Table 6

Wijma parent material is younger with an age of Water physical and chemical parameters from Hadash and Wijma springs.
10.193 ± 30 yr cal BP. The dated samples from Hadash and Hubra Hadash spring Wijma spring
were the most cemented soils (pore space ~15% in Hadash and < 5%
in Hubra) whereas the soils from Wijma are more weakely cemented pH 7.9 ± 0.02 7.5 ± 0.1
E.C. (μS/cm) 830 ± 8 536 ± 6
(~30% of pore space) as shown in Fig. 6a, e, f. An extensive discussion
D.O. (mg·L−1) 4.31 ± 0.04 5.42 ± 0.05
of the age results is provided in Section 5.3. ORP (mV) −7.7 ± 5 34 ± 5
Temperature (°C) 22.00 ± 0.1 20.00 ± 0.1
4.3. Spring water analyses and precipitation index Hardness (–HCO3, mg·L−1) 88 ± 5 71 ± 5
Cl (mg·L−1) 28.0 ± 0.1 33.0 ± 0.1
B (mg·L−1) 0.084 ± 0.001 0.054 ± 0.001
Table 6 shows the results of the water analyses. They exhibit fresh
Na (mg·L−1) 28.63 ± 0.02 32.14 ± 0.02
water quality with low hardness for both springs (71–88 ± 5 mg l−1). Mg (mg·L−1) 4.56 ± 0.02 3.65 ± 0.02
The spring water of Hadash is more alkaline (pH of 7.9) than that of Ca (mg·L−1) 10.43 ± 0.02 6.70 ± 0.02
Wijma (pH of 7.5). The Hadash spring has a higher content in Ca2+ and Al ((mg·L−1) 0.04 ± 0.01 0.25 ± 0.01
P (mg·L−1) 0.05 ± 0.001 0.04 ± 0.001
Mg2+, which explains the higher alkalinity compared to the spring in
K (mg·L−1) 1.017 ± 0.001 0.369 ± 0.001
Wijma. The annual precipitation in Fig. 7a shows a precipitation peak Sr (mg·L−1) 0.079 ± 0.002 0.064 ± 0.002
for the three meteorological stations in 1999 and then a drop in the Mn (mg·L−1) 0.0108 ± 0.0003 0.0150 ± 0.0003
following 19 years. Similarly, the precipitation index was mostly ne- Fe (mg·L−1) 0.193 ± 0.003 0.303 ± 0.003
gative after 1999 (Fig. 7b).
4.4. Soil thin section analyses microsparite of brownish gray color, while in some cases clots of micrite
are observed (Fig. 6a see the arrows). Voids and cracks are also present
The Hadash calcrete is a matrix-supported wackestone with (~15%) (Fig. 6a). In several cases, the cracks are wrinkled. There are
10
600 3 Saiq
Saiq Samail
(a) Samail
(b)
2.25 Seeb
Seeb
Annual precipitation (mm)
450
1.5
PI
300 0.75
0
150
-0.75
0 -1.5
1983 1988 1993 1998 2003 2008 2013 2018 1983 1988 1993 1998 2003 2008 2013 2018
Year Year
Fig. 7. (a) Rainfall data from Saiq, Samail and Seeb meteorological stations (data from 1983 to 1990 were rather fragmented), (b) Precipitation index (PI), positive
numbers showed “wet” years with higher average rainfall compared to the average rainfall of the period of available data in the meteorological station (see at Section
3.4), negative values showed years of drought.
sub-rounded shapes filled with sparry calcite, which are engulfed by a case in our study area (Buerkert et al., 2010). The hue in Munsell color
thin micrite layer (Fig. 6b). The sparry calcite-filled voids are showing scale is reddish-brown to dark reddish-weak red in Hadash and Wijma
anhedral equant calcite substituting xylem cells (Microcodium) soils respectively, probably due to ferrihydrite and lepidocrocite for-
(Fig. 6b). Small coal fragments are present. Several fragments of os- mation which are denoting oxidizing conditions. Ferrihydrite and le-
tracod shells (Fig. 6c, d) were identified. While some of them show pidocrocite have not been identified by XRD in the bulk soils, because
ornamentation (wavy surface) (Fig. 6c), they usually exhibit thin and to their poor crystallization and their few diffraction peaks, which may
smooth shells and more than two shell layers. All of them display the also overlap with phyllosilicate peaks (Jones and Saleh, 1986). The low
characteristic radial sweeping extinction (Fig. 6d). In one case, a part of aggregate stability succeeds the loss of organic matter after soil aban-
the recurved shell is preserved. Extraclast material of some unidentified donment in our study area. This obvious decrease of the aggregate
layered clasts (possibly slate or shale) is present. stability is in a linear contradiction to the soil response in other climatic
Wijma calcrete more fractured with a greater abundance of voids conditions such as continental-wet and semi-arid climates (Dunjó et al.,
(~30%) (Fig. 6e). It is a wackestone of micrite (dark brown) to mi- 2003; Stamati et al., 2013). It is generally accepted that the cultivated
crosparite. Extraclasts are made of some layered lithoclasts, possibly soils show higher C turnover related to tillage and the type of organic C
slate and shale (Fig. 6e). Calcified xylem cells with sparry anhedral addition (Six et al., 2000). On the other hand, the C turnover is low in
calcite (Microcodium) are coated by a thin rim of microsparite (Fig. 6f), non-cultivated soils, thus, preserving C in soil. However, this is not the
similar to what we found in the Hadash area. We also identified clastic case in the places of the present study. Taking the average C content
material of coarse limestone and small fragments of charcoal. Frag- (45 g kg−1 in the beginning of a cropping cycle) from four cultivated
ments of ostracods were identified by their radial sweeping extinction mountainous oases in Oman, presented by Al-Rawahi et al. (2014), we
(Fig. 6g, h). can calculate a loss of at least 88% of C compared to the average C
The Hubra calcrete is a packstone with only a few voids (< 5%), content in our study area. The loss of low density labile organic matter
while most of the space has been filled with micrite (Fig. 6i, j). Some is particularly affected by higher temperature in arid climates (Wagai
peloids of micrite were observed, while clotted micrites are common et al., 2008). Thus, the soil abandonment in the arid areas clearly de-
with intergrown spar. The clastic material consists of quartz and pla- teriorated the soils due to immediate loss of organic carbon. It has been
gioclase grains with angular to subrounded shapes (Fig. 6j). Less dis- widely observed that the loss of soil organic carbon is directly related to
cernible bioclastic fragments stem from ostracods and red algae (Fig. 6k aggregate disintegration in soil (Six et al., 2000; Yu et al., 2012). The
and l, respectively). Wrinkled cracks are absent. The grains are sepa- current condition of the soil quality has derived from the absence of
rated by micrite and displacive sparite. organic carbon addition (soil terraces abandonment) and the con-
comitant rather rapid organic carbon loss due to the arid conditions.
5. Discussion
5.2. Geochemical mineralogical and morphological evidence of regolith
5.1. Soil quality and evolution in the mountainous abandoned terraces provenance
The depths of the studied soil terraces are similar to that of others The presence and choices of unconsolidated sediments which could
around the globe, including, for example, from circum-Mediterranean be a potential constituent of a composite soil in the Sultanate of Oman
regions where terrace soil depths range from 30 to 80 cm (Zdruli et al., are very limited. Field work and analyses of satellite imagery, using
2014; Moraetis et al., 2015). The profiles in the lower terraces (H3 and Google Earth, showed that only few possible sources of fine-grained
W3 in Fig. 4a, b) display lower soil depths which could reflect soil material exist as potential components for anthropogenic pedogenesis.
erosion in terraces of lower elevation. The low TOC content Farmers in Oman are collecting and transporting regolith for artificial
(2.1–1.1 g kg−1) shows that terrace abandonment has negatively af- terrace soil creation and preservation even today (facts orally declared
fected soil quality and rendered the area prone to erosion. Similar by locals). Our investigations on the origin of the unconsolidated ma-
mountain oases in Oman, but with cultivated soil terraces, contain as- terials show that soils in Wijma and Hadash could not have been formed
tonishingly higher TOC content from 16 to 37 g kg−1 in a depth of in-situ because of their different geochemical characteristics when
15–100 cm (Luedeling et al., 2005). Oxidation of organic matter and compared with their bedrocks. The fact that the soils of Hadash re-
rapid increase of carbon turnover is common in hot, arid areas as is the semble those of Wijma despite different bedrock signatures shows that
11
the soils may have a common parent material (regolith) and shared a geochemistry and mineralogy of the Hadash and Wijma soils. For ex-
common management practice of artificial pedogenesis. ample, mixing lacustrine deposits with sediments from sand dunes that
The geochemical analysis reveals similarities between the Hadash are scattered between Wadi Mistal and Wadi Sahtan may also be con-
and Wijma soils which display similar contents of SiO2 and CaO. Both sidered as a possible source of parent material. However, the absence of
soils align with the PC1, although SiO2 and CaO percentages usually eolian quartz grains in thin sections and the high amount of calcareous
provide for a high data variability. The PC2 shows that soils from components does not support this idea. Only the paleosol sample from
Hadash and Wijma are grouped together into positive PC2 loading due Hubra reflects a possible mixture of eolian and fluvial sediments with
to a slightly higher content of P2O5, SO2 and Fe2O3. The amount of both angular and rounded grain shapes (Fig. 6i, k).
P2O5, SO2 and Fe2O3 in the studied soils resembles the cultivated soils
from Wadi Abyad and Wadi Sahtan. Especially the P2O5 content could
be a relict resulting from intense fertilization in the past. The Fe2O3 5.3. Radiocarbon and stable isotope results, validation and limitations
contents of the Hadash and Wijma soils are higher compared to all other
sediments and soils, which is possibly related to silicate oxidation due Dating of soil terraces is a challenging task. There are two kind of
to the frequent alternations of wet and dry periods. The soils referred to ages which can be obtained, the age of the construction of the terraces
as “other soil terraces” (Wadi Abyad and Wadi Sahtan, Hubra, As Sab) and the age of the infilling material. The first can be determined by
in Tables 3 and 5 as well as in Fig. 5a exhibit different geochemical optically stimulated luminescence (OSL), taking into account several
signatures in the PC analyses. The higher content of Al2O3 in the Ha- limitations (Kinnaird et al., 2017). In addition, 14C in charcoal or other
dash and Wijma soils, compared to all other soil terraces may be largely organic debris can also yield the age of terrace construction assuming
related to clay minerals (e.g., illite and vermiculite) which indicate a that organic material was introduced during infilling of the terraces
high degree of weathering. Although we did not carry out analyses to (Pietsch and Mabit, 2012). In the present study, we have relatively
identify mixed illite-vermiculite layers in the clays, which are indicative shallow soil profiles and lack of charcoal particles (except in micro-
of intense weathering, the presence of illite and vermiculite in the man- scopic size) or other visible organic debris. Thus, attempting to produce
made soil terraces of Hadash and Wijma indicates a high degree of an absolute time constraint of the terrace construction we were re-
weathering (Yin et al., 2018). stricted to date the soil material. The limitations of dating caliche have
The thin section analysis further contributes to the argument of an been presented in several elaborated studies (Chen and Polach, 1986;
off-site development of the soils in Hadash and Wijma. The soils show Williams and Polach, 1971). The prevailing trends regarding 14C use in
the presence of prevailing calcite (micrite and microsparite) and clay paleosols support the idea that ages of paleosols which have been af-
minerals and absence of regolith. Dissolution and weathering of bed- fected by calcite precipitation can be either too old or too young. This
rock material could have contributed to the soils, but there is no clear can be due to the different sources of carbonates in soil (Chen and
evidence of the bedrocks within the studied soils. As it is expected, only Polach, 1986). For example, a “limestone dilution effect” may derive
the sample from the old Wijma Village, which was collected as an ex- from the mixing of the actual sample 14C with 14C (“dead”) from
ample of in-situ soil development, shows clear geochemical affinities to limestone that is much older. The most reliable method to check
the subjacent bedrock (e.g., similar SiO2 percentages in soil and bed- whether a soil carbonate age is accurate is to compare the soil carbo-
rock, Table 2; similar mineralogy, except the dolomite absence from the nate dates with dates from organic carbon residues (e.g. charcoal) or
in-situ soil, Table 4). The studied soils show a similar regolith origin with other independently determined ages (e.g. archaeological evi-
that is most likely related to artificial in-filling of the terraces with dence). In such studies, in arid pedogenic carbonates, dating results
regolith. Moreover, the soil profiles show no transitional layers such as varied from 490 to 6950 a (average of 3460 yrs) with pedogenic car-
altered bedrock (in the sense of saprolite) as has been observed in other bonates to be older due to “dead” carbon incorporation (Williams and
artificial terraces with similar bedrock (see “K4 soil samples” in Polach, 1971). There are also other cases where authigenic carbonate
Moraetis et al. (2015)). segregation (pedogenic carbonates) may result in younger ages com-
The most convincing evidence that denotes a composite origin of the pare to the soil either from later pedogenic carbonates development
soils in Hadash and Wijma, is the presence of various ostracod frag- which postdates soil formation, and/or modern 14C incorporation from
ments (Fig. 6c, d, g, h). The presence of these microscopic bioclasts frequent rainfalls (atmospheric) (Chen and Polach, 1986; Pustovoytov
illustrates that the composite soil contains a portion of non-con- et al., 2007). Thus, we acknowledge the limitations of our radiocarbon
solidated material from a lacustrine environment. The presence of os- dating results and rely on multiple lines of evidence, including thin
tracods has not yet been reported from other artificial soil terraces in a section analyses, stable isotope results and evidence past climatic
mountainous setting in the Sultanate of Oman or elsewhere. Similarly, conditions.
archaeological excavations in soils from Balad Seet in the Jabal Akhdar The 14C ages for Hadash and Wijma soils (13.887 ± 40 yr cal BP
area (37 and 9 km distance in a straight line from Hadash and Wijma, and 10.193 ± 30 yr cal BP for Hadash and Wijma, respectively) pre-
respectively) have not shown any marine or lacustrine bioclasts (Nagieb date the onset of agriculture in this area. Our ages are based on dating
et al., 2004). Thus, we suggest that the soils of the terraces in Hadash calcretes. The thin sections show Microcodium with a thin rim of mi-
and Wijma contain human-transported material from a location of la- crosparite around it (in both sampling sites). The rim bridges
custrine deposits with ostracods. Microcodium to the matrix, indicating that the calcrete development
To identify possible mining localities of the lacustrine sediments, we happened in-situ (Fig. 6a, b, e, f). Identical layered micritic material
considered and studied the lake deposits of Wadi Mistal for Hadash and and Microcodium-like structures have been identified in calcretes within
those of Wadi Sahtan for Wijma. The samples from the two places Holocene terra rossa1-filled fissures in the Jabal Akhdar area (less than
display a similar CaO content as the soil samples (Table 3) and a cri- 10 km SSW from Hadash) (Khalaf and Al-Zamel, 2016). Since Micro-
tically similar fine texture (Hoffmann et al., 2015). However, as the codium (significant part of calcrete) was formed in-situ, the calcite age
PCA and the mineralogical analyses show they were geochemically and should have been younger than the initiation of agriculture in Oman
mineralogically different compared to the soils of Hadash and Wijma. (approximately 5–3 yr BP) (Nagieb et al., 2004). Obviously, the soil age
We have no microscope analysis for the lacustrine deposits to decide if is much older, according to our results. We consider it likely, that in-situ
the ostracods which are present in those potential sources of soil parent
material are similar to those in our studied soils. Thus, we cannot rule 1
Definitions of terra rossa: (1) Chromic luvisol (FAO 2006), (2) A reddish-
out the lacustrine deposits of Wadi Mistal and Wadi Sahtan as the brown residual soil found as a mantle over limestone bedrock, typically in the
parent material of the soil composite, since a mixture of lacustrine se- karst areas around the Adriatic Sea, under conditions of Mediterranean-type
diments with other materials could have impacted the final climate (Jackson 1997).
12
calcrete formation was created relatively late by neomorphism of in- Cerling, 1995) that could have contributed to young ages by in-
itial, early fine-grained calcite precipitates. The presence of micro- corporating atmospheric carbon. However, the age that we have cal-
sparite with remnants of micrite, especially in the Hadash soil, where culated predates initiation of agriculture in Oman, and the aridity of the
the cementation was intense, shows an evolution of diagenesis with a climate during the past 6000 years are not favoring the incorporation of
constant supply of calcium and carbonate. The in-situ calcrete pre- modern 14C. Therefore, the ages of the present study are definitely older
cipitation probably progressed from the upper towards the lower soil than the onset of agriculture in Oman, with stable isotopes analysis
horizons as the CaO concentration increases in the lower horizons indicating similar age carbonates elsewhere in Oman, with unlikely
(Table 3, H1 and H2 profiles). Thus, we suggest that the calcrete was influence of a younger age trend due to incorporation of atmospheric
supplied with calcium and carbonate that was made available by dis- carbon (plants and irrigation water).
solution of calcite from the upper soil horizons, and the age of calcrete Using the 14C results as a relative dating for comparison to the
is recording mainly the age of the initial fine-grained calcite crystals calcrete development in the studied samples, also offers a valuable ar-
which was used as terrace infill. However, the carbonates added by late gument on the 14C origin. The fact that Wijma, which exhibits a very
irrigation may not be neglected, but those would have provided a low calcrete development compared to Hadash, has a younger age
younger age due to possible modern 14C incorporation. It is worth (10.193 ± 30 ka BP and 13.887 ± 40 yr cal BP, respectively) likely
mentioning that in the thin sections we found no extraclasts of Meso- indicates that the more calcite precipitation proceeded the more
zoic and Paleozoic carbonates that could have caused a much earlier pedogenic carbon is incorporated from the brought-in calcite. We sug-
age in our soils except in the Wijma sample, which actually shows a gest that the Hadash age represents a lower age boundary which is
younger age than the Hadash sample. Thus, combining microscopic corroborated by the Hubra sample. The Hubra sample was selected as a
results with dating results we are confident that the estimated age is not typical and very well developed paleosol calcrete lacking Microcodium
deviating from the actual age of the soil. that could denote incorporation of “young” (e.g. atmospheric) carbon.
Another line of evidence to verify the dating results, is the use of the Cementation in Hubra is more intense (absence of open cracks and
stable isotopes and their comparison with other published results on displacive cement) compared to the sites of Hadash and Wijma and
recent calcite depositions. The δ13C content is enriched in all samples, shows a similar age to Hadash (with a maximum boundary of age the
especially in the Hubra soil. Fig. 8 depicts the different ages of tra- Hubra sample). This leads to the argument that the more calcrete de-
vertines and the water stable isotopes from the Jabal Akhdar area (Falk veloped the more pedogenic carbon was incorporated. We suggest that
et al., 2016; Matter et al., 2005; Mervine et al., 2014). The stable iso- the age of the calcrete does not reflect the cementation age but largely
topes of water samples from the Jabal Akdhar area are quite close in the age of the initial sediment used for anthropogenic pedogenesis on
concentration to our samples. Our data exhibit a stable isotope content the terraces, taking also into account a possible younger trend.
similar to that of samples with an age between 6000 and 18,000 yr BP Furthermore, the following arguments show that the presented soil
(empty circles in Fig. 8). Mervine et al. (2014) reported that generally ages reflect an effective carbon pool related to sediments deposited
older travertines (> 1000 years) are enriched in stable isotopes and the during an arid period in Oman. The climatic conditions during the Late
samples which are actually close to the current water signature are Pleistocene to Early Holocene including the Younger Dryas
those of an age between 6000 and 18,000 yr BP (empty circles in (12,900–11,700 a BP) are of great interest especially in this part of the
Fig. 8). The recent travertines, (< 1000 years) such as the samples from world where particular geomorphological features, e.g., certain glaci-
surface film of calcite in Oman alkaline springs (Falk et al., 2016) and genic deposits, are absent. Soil core samples from the Jabal Akhdar area
modern travertine deposits from Misht and Qafeefah (Mervine et al., show silt accumulation and a CaCO3 content of 40% with an age of
2014) are showing a trend of depletion in both δ13C and δ18O which is 13.200–11.400 BP (Urban and Buerkert, 2009). Paleoclimatic studies
attributed to kinetic fractionation of present-day deposition (Falk et al., from cave stalagmites in the Hooti and Qunf caves of Oman and the
2016). Thus, our stable isotopes share similar trends with other pub- Socotra Cave of Yemen clearly verified a period of dry conditions prior
lished samples from travertine within an age range between 6 and 10.6 ka BP within the Intertropical Convergence Zone (ITCZ) south of
18 ka BP (Mervine et al., 2014). On the other hand, the δ13C content of Oman (Fleitmann et al., 2011). In addition, Fuchs and Buerkert (2008)
our samples is also falling within the range of C4 plants (Quade and indicated a period of low rainfall with increased sedimentation of clay
Fig. 8. Stable isotopes δ13C‰ and δ18O‰ from Hadash (H), Wijma (W) and Hubra (Hu). The groundwater data (*) were from Matter et al. (2005), the travertines of
modern to 43 ka BP of age (○, ♢, ✕) were from Mervine et al. (2014), the recent film of travertine (+) were from Falk et al. (2016). Rain was from Matter et al.
(2005) and references therein.
13
and silt along with high contents of CaCO3 between 19.5 and by the action of the limited rhizosphere even during arid periods by
12.5 ka BP. The same authors argued that especially for the period exchange of H ± with nutrients including Ca2+ (McConnaughey and
between 12.5 and 10.5 ka BP the rainfall was even less with infrequent Whelan, 1997). Fluids rich in Ca2+ reached lower horizons where they
rains that probably coincide with the Younger Dryas event in Europe. can precipitate carbonate. Indeed, Abed et al. (2013) showed that
Thus, accepting the carbon age that we estimated in the present study, during the first rains in Oman nitrification quickly progresses into de-
the material’s origin was probably related to small mountain ponds that nitrification in the desert coinciding with a high loss of N2O (in-
evaporated and forced fine calcite precipitation during dry periods and complete denitrification). The subsequent denitrification especially at
strong erosion during erratic rainfalls. These pond materials were lower soil horizons and at relative anoxic conditions, promotes a final
transported directly from the mountains of the Jabal Akhdar area into pH rise with contemporaneous calcite precipitation (Hamdan et al.,
nearby wadis and probably represent a major constituent of the infill of 2017). Denitrification can also be driven by abiotic processes such as
the soil terraces, for at least the early stages of terrace construction photochemical alteration to N2O (Georgiou et al., 2015) and/or by
(maybe 700–1000 years BP as locals declare). Today, such material oxidation of Fe and the production of peroxide (Fenton’s reaction) in
only occurs in a few wadis in Oman where perennial flow exists and desert soils (Nyamunda et al., 2013). Similarly, to the conditions which
they are in large distance from our studied sites. In Wadi Mistal and favor Microcodium formation, the alternations of short wet periods with
Wadi Sahtan the only fine-grained sediments are the recent dam de- long dry periods in the arid areas are enhancing complex biotic pro-
posits of Wadi Mistal. However, this material differs from the soil ma- cesses that are mobilizing and re-precipitating calcite. We propose that
terial of Hadash. The sediments contemporaneous with the Younger a combination of poorly fertilized soils in fallow periods along with the
Dryas are the parent material of the soil terrace infills, and they have complex biotic and abiotic conditions during the alternations of wet
been widely eroded and are largely missing at today’s surface. and dry periods were the driving factors of the brought-in fine calcite
dissolution and re-precipitation of calcrete. This argument is in agree-
5.4. Calcrete formation, environmental implications and historical facts ment with the 14C origin in our results which largely reflects the
brought-in fine calcite.
Calcrete formation in soils has a remarkable deteriorating effect on The calcrete formation is favored in the arid areas and soil man-
the productivity of the soils. The cemented soil horizon is prohibiting agement of periodically cultivated soils which were abandoned in ex-
water infiltration and soil aeration with the result of water stagnation treme dry periods, as during the last 10–15 years in Oman (Fig. 7a, b) is
and plant rot. It also hinders root growth. The effect of calcrete was also enlarging the impact. Alternating wet/dry conditions have been ob-
mentioned by local people who witnessed plant rot in the Village of served by others (Kwarteng et al., 2009), and could be related to El
Hadash where the most extensive calcrete formation was observed. The Nino-Southern Oscillation (Burns et al., 2002). A dry period is currently
absence of calcrete from other soil terraces in the mountainous areas of underway, and soil terraces are under stress in several areas of Oman as
Oman may indicate differences in the local environmental conditions in our study area. The characteristic decline of agriculture in our study
such as lithology and irrigation water quality. However, the published area is shown in Fig. 9. The cultivation of perennial crops and palm
data from similar soil terraces in a mountainous area in Oman showed trees is absent today in Wijma (Fig. 9a, b). This decline has also been
similar soil lithology regarding the CaCO3 content. The irrigation water recorded in historical archives of Google Earth (Fig. 9c–e) from 2004
quality as presented in Luedeling et al. (2005), (their Table 4) was si- until 2018. Alternating wet/dry periods have probably transformed the
milar to the samples tested in our study area and with an EC from 500 area from verdant to denuded several times during the cultivation
to 800 μS cm−1, while the Ca content was much lower in our study area history of the terraces. These land cover oscillations along with their
(10–6 mg l−1) compared to the previous study (114 mg l−1). Our pH adverse effects have not been observed in other mountainous soil ter-
values were slightly more acidic (7.5–7.9) compared to 8.2 in the ones races such as in Balad Seet (Buerkert et al., 2010; Nagieb et al., 2004).
of the previous study. The slightly more acidic pH could have been We suggest that this results from the lower elevation of the Balad Seet
enhancing calcite dissolution and redeposition in our study area. (995 m) compared to our study area (Hadash elevation: 1472 m and
However, we point out that another critical factor for the precipitation Wijma elevation: 1262 m). A fresh water spring at lower elevations
of calcrete was of biological origin. The identification of Microcodium generally drains a much larger watershed compared to those at higher
(Fig. 5b) verified the strong influence of the biological factors especially elevations. Thus, the agricultural areas at lower elevations were not
in the periods of no cultivation and the prevalence of Ca2+ as the only seriously affected by the recent 10–15 years drought, which reduced the
available nutrient. Kosir (2004) supported the idea that cell replace- water supply at higher elevations. Other factors such as the significant
ment by calcite is due to roots’ unsuccessful attraction of nutrients other socio-economic changes in Oman in the course of petroleum production
that Ca2+. Thus, calcified roots could show periods when soils were during the last five decades may have also caused an abandonment of
poor in nutrients and probably similar to todays’ conditions of soil terraces although climatic factors seem to be the main cause of aban-
abandonment and lack of fertilization. On the other hand, a study by donment.
Kabanov et al. (2008), contradicts the rhizosphere related origin of
Microcodium and indicates that saprotrophic microorganisms are the 6. Conclusions
driving factor behind calcite precipitation. Saprotrophic microorgan-
isms (e.g. fungi) in arid environments decompose biomass in a pulse We present results of soil quality and soil provenance analyses for
like mode during and shortly after the wet periods (Collins et al., 2008). two areas with abandoned soil terraces. Our findings show that the
The development of saprotrophic microorganisms is favoured in arid, abandoned soil terraces exhibit very low TOC and low soil aggregate
carbonate-rich soil and they can dissolve calcite with exuded oxalic acid stability. Longer lasting processes, like calcrete formation, further de-
and subsequently re-precipitate it after the pulse like decomposition teriorated the soil quality, particularly in terraces consisting of a unique
(Kabanov et al., 2008). In both above interpretations of Microcodium composite soil. Our results indicate that the regolith in these composite
origin, the existence of short wet periods and prolonged arid periods are soils, which shows a lacustrine origin (ostracods), arrived via human
triggers of calcite re-precipitation in carbonate rich soil. transport. The age of the regolith matches well with a period of dryness
In addition, it is well known that NH4 ± builds up in desert soils and subsequent calcite deposition, such as the Younger Dryas. The
during the dry periods (Stursova et al., 2006). After prolonged arid identification of human-transported material represents a major in-
intervals, the first rain events may also promote excess NH4 ± ni- vestment of labor and offers new evidence regarding construction of
trification and acidification of soil in a manner similar to excessive terraces in a mountainous, arid and tropical locale.
NH4+ fertilization (Mills et al., 2011). This acidification can mobilize We suspect that terrace abandonment recurred several times in the
Ca2+ through calcite dissolution. Acidification also has been observed agricultural history of the study area and played a significant role in
14
(a) (b)
(c) (d) (e)
200 m 200 m
200 m
Fig. 9. (a) Wijma in the 1990s (photograph was provided by Mazin Sulaim Sunan Al-Abri), (b) Wijma in 2019, observe the cluster of houses for ease of comparison.
Google Earth image from the historical view tool, images correspond to (c) 11/1/2004, (d) 28/2/2009, (e) 21/2/2018.
calcrete formation and diminishing soil quality. Most recently, the Belize. Geogr. Rev. 92, 372–397.
contemporary abandonment of agricultural terraces represents a sig- Béchennec, F., Roger, J., Le Métour, J., Wyns, R., 1992. Geological map of Seeb, sheet NF
40-03, scale 1:250,000, with Explanatory Notes: Directorate General of Minerals,
nificant loss of arable land and an important cultural heritage resource Oman Ministry of Petroleum and Minerals.
suitable for eco-tourism. Béchennec, F., Le Métour, J., Platel, J.P., Roger, J., 1993. Geological map of the Sultanate
The presented combination of geochemical and mineralogical of Oman (GIS version), 1:250,000. In Explanatory Notes, Ministry of Petroleum and
Minerals, Directorate General of Minerals.
techniques represents a significant tool for unravelling the agricultural Beurrier, M., Béchennec, F., Hutin, G., Rabu, D., 1986b. Geological map of As Suwayq,
history of a study area. Moreover, we hope to pursue archaeological sheet NF 40-03A, scale 1:100,000, with Explanatory notes: Directorate General of
investigations in the future to significantly contribute to the under- Minerals, Oman Ministry of Petroleum and Minerals.
Bevan, A., Conolly, J., 2011. Terraced fields and mediterranean landscape structure: an
standing of anthropogenic pedogenesis in Oman. analytical case study from Antikythera. Greece. Ecol. Model. 222, 1303–1314.
Bevan, A., Conolly, J., Colledge, S., Frederick, C., Palmer, C., Siddall, R., Stellatou, A.,
Declaration of Competing Interest 2013. The long-term ecology of agricultural terraces and enclosed fields from anti-
kythera. Greece. Hum. Ecol. 41, 255–272.
Bouyoucos, J.G., 1962. Hydrometer method improved for making particle size analyses of
The authors declare that they have no known competing financial soils. Agron. J. 54, 464–465.
interests or personal relationships that could have appeared to influ- Burns, S.J., Fleitmann, D., Mudelsee, M., Neff, U., Matter, A., Mangini, A., 2002. A 780-
ence the work reported in this paper. year annually resolved record of Indian Ocean monsoon precipitation from a spe-
leothem from south Oman. J. Geophys. Res. 107, 1–9.
Buerkert, A., Luedeling, E., Dickhoefer, U., Lohrer, K., Mershen, B., Schaeper, W., Nagieb,
Acknowledgements M., Schlecht, E., 2010. Prospects of mountain ecotourism in oman: the example of as
sawjarah on al jabal al akhdar. J. Ecotourism 9, 104–116.
Chen, Y., Polach, H.A., 1986. Validity of 14C ages of carbonate in sediments. Radiocarbon
We are thankful to Hamdan Al-Zindi for the thin section production 28, 464–472.
and the technicians of the Earth Sciences Department at Sultan Qaboos Charbonnier, J., 2017. The Genesis of Oases in Southeast Arabia: Rethinking Current
University for their assistance. We are also thankful to the Center of Theories and Models, in Oases and Globalization. Springer, Cham, pp. 53–72.
Collins, L.S., Sinsabaugh, L.R., Crenshaw, C., Green, L., Porras-Alfaro, A., Stursova, M.,
Analytical Applied Unit (CAARU) at Sultan Qaboos University, for XRF Zeglin, H.L., 2008. Pulse dynamics and microbial processes in aridland ecosystems. J.
analysis and mineralogical analysis. Ecol. 96, 413–420.
Dunjó, G., Pardini, G., Gispert, M., 2003. Land use change effects on abandoned terraced
soils in a mediterranean catchment, NE Spain. Catena 52, 23–37.
References Faccini, F., Gabellieri, N., Paliaga, G., Piana, P., Angelini, S., Coratza, P., 2018.
Geoheritage map of the Portofino natural park (Italy). J. Maps 14 (2), 87–96.
Abdalla, O., Al-Abri, Y.B.R., 2011. Groundwater recharge in arid areas induced by tro- Falk, S.E., Guo, W., Paukert, N.A., Matter, M.J., Mervine, M.E., Kelemen, B.P., 2016.
pical cyclones: lessons learned from Gonu 2007 in Sultanate of Oman. Environ. Earth Controls on the stable isotope compositions of travertine from hyperalkaline springs
Sci. 63, 229–239. in Oman: insights from clumped isotope measurements. Geochim. Cosmochim. A
Abed, M.M.R., Lam, P., de Beer, D., Stief, P., 2013. High rates of denitrification and 192, 1–28.
nitrous oxide emission in arid biological soil crusts from the Sultanate of Oman. ISME FAO, 2006. World Reference Base for Soil Resources, World Soil Resources Reports No.
J. 7, 1862–1875. 103, second ed. Food and Agriculture Organization of the United Nations FAO, Rome.
Al-Rawahi, N.M., Brinkmann, K., Schlecht, E., Buerkert, A., 2014. Effects of changing Fleitmann, D., Burns, S.J., Pekala, M., Mangini, A., Al-Subbary, A., Al-Aowah, M.,
water availability on land use in irrigated mountain oases of Al Jabal Al Akhdar, Kramers, J., Matter, A., 2011. Holocene and pleistocene pluvial periods in Yemen,
northern Oman. Erde 145 (4), 197–211. southern Arabia. Quat. Sci. Rev. 30, 783–787.
Beach, T., Luzzadder-Beach, S., Dunning, N., Hageman, J., Lohse, J., 2002. Upland Fuchs, M., Buerkert, A., 2008. A 20 ka sediment record from the Hajar Mountain range in
agriculture in the Maya Lowlands: ancient Maya soil conservation in North-western N-Oman, and its implication for detecting arid–humid periods on the southeastern
15
Arabian Peninsula. Earth Planet. Sci. Lett. 265, 546–558. Mills, T.C., Morrison, M.J., Goldhaber, B.M., Ellefsen, J.K., 2011. Chromium(VI) gen-
Georgiou, D.C., Sun, J.H., McKay, P.C., Grintzalis, K., Papapostolou, I., Zisimopoulos, D., eration in vadose zone soils and alluvial sediments of the southwestern Sacramento
Panagiotidis, K., Zhang, G., Koutsopoulou, E., Christidis, E.G., Margiolaki, I., 2015. Valley, California: a potential source of geogenic Cr(VI) to groundwater. Appl.
Evidence for photochemical production of reactive oxygen species in desert soils. Nat. Geochem. 26, 1488–1501.
Commun. 6, 7100. https://doi.org/10.1038/ncomms8100. Moraetis, D., Nikolaidis, N.P., Paranychianakis, N., Rousseva, S., Kercheva, M., Nenov,
von Grobe, A., Hagke, C., Littke, R., Dunkl, I., Wübbeler, F., Muchez, P., Urai, J.L., 2019. M., 2015. Soil genesis, evolution, and properties in the koiliaris river watershed-
Tectono-thermal evolution of Oman’s Mesozoic passive continental margin under the critical zone observatory. J. Soil Sediment. 15, 347–364.
obducting semail ophiolite: a case study of Jebel Akhdar, Oman. Solid Earth 10, Moraetis, D., Mattern, F., Scharf, A., Frijia, G., Kusky, M.T., Yuan, Y., El-Hussain, I., 2018.
149–175. https://doi.org/10.5194/se-10-149-2019. Neogene to quaternary uplift history along the passive margin of the northeastern
Hamdan, N., Kavazanjian Jr., E., Rittmann, E.B., Karatas, I., 2017. Carbonate mineral Arabian Peninsula, eastern Al Hajar Mountains, Oman. Quat. Res. 90, 418–434.
precipitation for soil improvement through microbial denitrification. Geomicrobiol. Nagieb, M., Siebert, S., Luedeling, E., Buerkert, A., Häser, J., 2004. Settlement history of a
J. 34, 139–146. mountain oasis in northern Oman-evidence from land-use and archeological studies.
Hansman, R.J., Ring, U., Thomson, S.N., den Brock, B., Stübner, K., 2017. Late Eocene Erde 135, 81–106.
uplift of the Al Hajar Mountains, Oman, supported by stratigraphic and low-tem- Novák, T.J., Incze, J., Spohn, M., Glina, B., Giani, L., 2014. Soil and vegetation trans-
perature thermochronology. Tectonics 36 (12), 3081–3109. https://doi.org/10. formation in abandoned vineyards of the Tokaj Nagy-Hill, Hungary. Catena 123,
1002/2017TC004672. 88–98.
Harrower, M.J., Nathan, S., 2018. Ancient water management in Southern Arabia: Nyamunda, C.B., Chigondo, F., Moyo, M., Guyo, U., Shumba, M., Nharingo, T., 2013.
creativity, resilience, and sustainability in Yemen and Oman. In: Sulas, F., Pikirayi, I. Hydrogen peroxide as an oxidant for organic reactions. J. Atoms Molecules 3, 23–44.
(Eds.), Water and Society from Ancient Times to the Present: Resilience, Decline, and Pietsch, D., Mabit, L., 2012. Terrace soils in the Yemen highlands: using physical, che-
Revival. Taylor & Francis/Routledge, Abingdon, UK, pp. 201–218. mical and radiometric data to assess their suitability for agriculture and their vul-
Hoffmann, G., Rupprechter, M., Rahn, M., Preusser, F., 2015. Fluvio-lacustrine deposits nerability to degradation. Geoderma 185–186, 48–60.
reveal precipitation pattern in SE Arabia during early MIS 3. Quatern. Int. 382, Pustovoytov, K., Schmidt, K., Parzinger, H., 2007. Radiocarbon dating of thin pedogenic
145–153. carbonate laminae from holocene archaeological sites. Holocene 17 (6), 835–843.
Jackson, J.A., 1997. Glossary of Geology, fourth ed. American Geological Institute, Quade, J., Cerling, T.E., 1995. Expansion of C4 grasses in the Late Miocene of North
Alexandria, Virginia, pp. 769. Pakistan: evidence from stable isotopes in paleosols. Palaeogeogr. Palaeocl. 115,
Jones, A.A., Saleh, A.M., 1986. Electron diffraction and the study of ferrihydrite coatings 91–116.
on kaolinite. Clay Miner. 21, 85–92. Rabu, D., Béchennec, F., Beurrier, M., Hutin, G., 1986. Geological map of Nakhl, sheet
Kabanov, P., Anadón, P., Krumbein, E.W., 2008. Microcodium: an extensive review and a NF40-3E, scale: 1:100,000, with Explanatory Notes: Directorate General of Minerals,
proposed non-rhizogenic biologically induced origin for its formation. Sediment. Oman Ministry of Petroleum and Minerals.
Geol. 205, 79–99. Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Grootes,
Kinnaird, T., Bolos, J., Turner, A., Turner, S., 2017. Optically-stimulated luminescence P.M., Guilderson, T.P., Haflidason, H., Hajdas, I., Hatt, C., Heaton, T.J., Hoffmann,
profiling and dating of historic agricultural terraces in Catalonia (Spain). J. Archaeol. D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M.,
Sci. 78, 66–77. Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M.,
Khalaf, I.F., Al-Zamel, A., 2016. Petrography, micromorphology and genesis of holocene van der Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves
pedogenic calcrete in Al-Jabal Al-Akhdar, Sultanate of Oman. Catena 147, 496–510. 0–50,000 years cal BP. Radiocarbon 55 (4), 1869–1887.
Kosir, A., 2004. Microcodium revisited: root calcification products of terrestrial plants on Shahalam, A.M., 2001. Review of Omani aflaj systems: an element of national water
carbonate-rich substrates. J. Sed. Res. 74, 845–857. resources, technology and economic developments. Sci. Technol. 6, 1–14.
Kosmas, C., Danalatos, N.G., Gerontidis, S.T., 2000a. The effect of land parameters on Siebert, S., Nagieb, M., Buerkert, A., 2007. Climate and irrigation water use of a mountain
vegetation performance and degree of erosion under mediterranean conditions. oasis in northern Oman. Agr. Water Manage. 89, 1–14.
Catena 40, 3–17. Six, J., Elliott, E.T., Paustian, K., 2000. Soil macroaggregate turnover and microaggregate
Kosmas, C., Gerintidis, S.T., Marathianou, M., 2000b. The effect of land use change on formation: a mechanism for carbon sequestration under no tillage agriculture. Short
soils and vegetation over various lithological formations on Lesvos (Greece). Catena Commun. Soil Biol. Biochem. 32, 2099–2103.
40, 51–68. Stamati, E.F., Nikolaidis, P.N., Schnoor, L.J., 2013. Modeling topsoil carbon sequestration
Koulouri, M., Giourga, Chr, 2007. Land abandonment and slope gradient as key factors of in two contrasting crop production to set-aside conversions with RothC–Calibration
soil erosion in mediterranean terraced lands. Catena 69, 274–281. issues and uncertainty analysis. Agr. Ecosyst. Environ. 165, 190–200.
Kotwicki, V., Al Sulaimani, Z., 2009. Climates of the Arabian Peninsula–past, present, Stursova, M., Crenshaw, L.C., Sinsabaugh, L.R., 2006. Microbial responses to long-term N
future. Int. J. Clim. Change Strategies Manage. 1, 297–310. deposition in a semiarid grassland. Microbial. Ecol. 51, 90–98.
Kwarteng, A.Y., Dorvlo, A.S., Kumar, V.T.G., 2009. Analysis of a 27-year rainfall data Tarolli, P., Preti, F., Romano, N., 2014. Terraced landscapes: from an old best practice to a
(1977–2003) in the Sultanate of Oman. Int. J. Climatol. 29 (4), 605–617. potential hazard for soil degradation due to land abandonment. Anthropocene 6,
Lasanta, T., Arnaíez, J., Oserín, M., Ortigosa, M.L., 2001. Marginal lands and erosion in 10–25.
terraced fields in the mediterranean mountains. A case study in the Camero Viejo Urban, B., Buerkert, A., 2009. Palaeoecological analysis of a late quaternary sediment
(Northwestern Iberian System, Spain). Mt. Res. Dev. 21, 69–76. profile in northern Oman. J. Arid Environ. 73, 296–305.
Londoño, A.C., 2008. Pattern and rate of erosion inferred from Inca agricultural terraces Varisco, D.M., 1991. The future of terrace farming in Yemen: a development dilemma.
in arid southern Peru. Geomorphology 99, 13–25. Agric. Hum. Values 8 (1–2), 166–172.
Luedeling, E., 2007. Sustainability of Mountain Oases in Oman. Effects of Agro- Varotto, M., Bonardi, L., Tarolli, P., 2019. World Terraced Landscapes: History,
Environmental Changes on Traditional Cropping Systems. PhD, Thesis. Faculty of Environment, Quality of Life. Springer, Cham, Switzerland.
Organic Agricultural Sciences of the University of Kassel, pp. 124–125. Vogel, H., 1988. Deterioration of a mountainous agro-ecosystem in the third world due to
Luedeling, E., Buerkert, A., 2008. Typology of oases in northern Oman based on Landsat emigration of rural labor. Mt. Res. Dev. 8 (4), 321–329.
and SRTM imagery and geological survey data. Remote Sens. Environ. 112, Wagai, R., Mayer, M.L., Kitayama, K., Knicker, H., 2008. Climate and parent material
1181–1195. controls on organic matter storage in surface soils: a three-pool, density-separation
Luedeling, E., Nagieb, M., Wichern, F., Brandt, M., Deurer, M., Buerkert, A., 2005. approach. Geoderma 147, 23–33.
Drainage, salt leaching and physico-chemical properties of irrigated man-made ter- Walkley, A., Black, A.I., 1934. An examination of the degtjareff method for determining
race soils in a mountain oasis of northern Oman. Geoderma 125, 273–285. soil organic matter, and a proposed modification of the chromic acid titration
Matter, J.M., Waber, H.N., Loew, S., Matter, A., 2005. Recharge areas and geochemical method. Soil Sci. 37, 29–38.
evolution of groundwater in an alluvial aquifer system in the Sultanate of Oman. Wilkinson, T., 1999. Settlement, agriculture highland Yemen: a preliminary statement. In
Hydrogeol. J. 14, 203–224. Proceedings of the Seminar for Arabian Studies, Vol. 29, Papers from the Thirty-
Mattern, F., Moraetis, D., Abbasi, I., Al Shukaili, B., Scharf, A., Claereboudt, M., Looker, Second meeting of the Seminar for Arabian Studies held in London, pp. 183–191.
E., Al Haddabi, N., Pracejus, B., 2018. Coastal dynamics of uplifted and emerged Late Williams, G.E., Polach, H.A., 1971. Radiocarbon dating of arid-zone calcareous paleosols.
Pleistocene near-shore coral patch reefs at Fins (eastern coastal Oman, Gulf of Oman). Geol. Soc. Am. Bull. 82, 3069–3086.
J. Afr. Earth Sci. 138, 192–200. Yin, K., Hong, H., Churchman, J.G., Li, Z., Fang, Q., 2018. Mixed-layer illite-vermiculite
McConnaughey, T.A., Whelan, J.F., 1997. Calcification generates protons for nutrient and as a paleoclimatic indicator in the Pleistocene red soil sediments in Jiujiang, southern
bicarbonate uptake. Earth Sci. Rev. 42, 95–117. China. Palaeogeogr. Palaeocl. 510, 140–151.
Mervine, M.E., Humphris, E.S., Sims, W.W.K., Kelemen, B.P., Jenkins, J.W., 2014. Yu, H., Ding, W., Luo, J., Geng, R., Cai, Z., 2012. Long-term application of organic manure
Carbonation rates of peridotite in the Samail Ophiolite, Sultanate of Oman, con- and mineral fertilizers on aggregation and aggregate-associated carbon in a sandy
strained through 14C dating and stable isotopes. Geochim. Cosmochim. A 126, loam soil. Soil Till. Res. 124, 170–177.
371–397. Zdruli, P., Calabrese, J., Ladisa, G., Otekhile, A., 2014. Impacts of land cover change on
Ministry of Regional Municipalities, Environment, and Water Resources/UNEP/ROWA/ soil quality of manmade soils cultivated with table grapes in the Apulia Region of
UNCCD, 2005. The National Plan of Action to Combat Desertification in the Sultanate south-eastern Italy. Catena 121, 13–21.
of Oman, pp. 65–80.
16

Geoderma 2020 Omanterraceagriculturefinal

Uploaded by

Copyright:

Available Formats

Geoderma 2020 Omanterraceagriculturefinal

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Geoderma 2020 Omanterraceagriculturefinal

Uploaded by

Copyright:

Available Formats

Geoderma 363 (2020) 114152

Contents lists available at ScienceDirect

Terrace agriculture in a mountainous arid environment – A study of soil T

ARTICLE INFO ABSTRACT

1. Introduction positive or negative influences on soil fertility and erosion (Kosmas

station. (Varian 710-ES from Varian Inc.) and Aquaread AP800.

H1 (0–5) AC 7.85 46 loam 2.5YR 5/3 4.90

Fractions (standard deviation) Η1 Η2 Η3 W1 W2 W3

Depth/horizon (5–16 cm)/C1 (5–9 cm)/AC

Depth/horizon (16–33 cm)/C2 (9–19 cm)/C1

Hadash Wijma Hubra

(a) (b) (c)

(d) (e) (f)

Possible soil parent rock

Other soil terraces

Hadash soil profiles exhibit extensively cemented blocky soil as shown

terraces of the latter have been constructed at a much steeper slope

Formation was met in the W3 profile of Wijma at a depth of 47 cm.

percentage (WSA) is < 42% in 8 out of the 10 the soil horizons of

Hadash and Wijma.

4.2. Results on bulk chemical analysis and dating

( ± 1.6%), respectively, which are also not statistically different

(P < 0.05). The bedrock in the Hadash area is dolomite with an

identical CaO abundance in the Hadash soil. However, the MgO

Bedrock in the old Wijam

derlying parent rock (Table 3).

Principal Component Analysis (PCA) of the 35 rocks and soils shows

variability. CaO and SiO2 concentrations display the highest variability

Soil-Hadash (wadi Mistal) ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Possible parent rocks

Other soil terraces

Example of insitu soil and parent rock

H1 38–57 cm Beta−452,901 12060–11815 13887 ± 40 BP −3.9 −2.2

(a) (b) (c) (d)

1000 µm 100 µm 200 µm 200 µm

(e) (f) (g) (h)

200 µm 200 µm 200 µm

(i) (j) (k) (l)

13.887 ± 40 yr cal BP and 13.186 ± 30 yr cal BP, respectively. The Table 6

(c) (d) (e)

You might also like