ReseaRch PaPeR Journal of Agricultural and Marine Sciences Vol. 21 (1): 33 – 46 DOI: http://dx.doi.org/10.24200/jams.vol21iss0pp33-46 Received 17 Sept 2015 Accepted 15 Feb 2016 Spatio-temporal dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping as an example Michael L. Deadman1*, Abdullah M. Al-Sadi1, Malik M. Al-Wardi2, Khalifa S.M. Al-Kiyumi3, W.M. Deadman4, and Fahad A. Al Said1 *1 Michael Deadman ( ) Sultan Qaboos University, College of Ag- ricultural and Marine Sciences, Department of Crop Sciences . Box 34, Al-Khod 123. Sultanate of Oman. email: mikedead@squ.edu.om. 2SQU, CAMS, Department of Soils, Water and Agricultural Engi- neering. 3 Ministry of Agriculture and Fisheries Wealth, P O Box 467, Muscat 113, Sultanate of Oman. 4Department of Archaeology, Durham University, South Road, Durham, DH1 3LE, UK Introduction Under the theoretical umbrella of land use/land cover change research, much attention has naturally focused on agro-forestry dynamics, including deforestation (Asner et al., 2005; Nepstad et al., 1999), agricultural expansion (Maeda et al., 2010) or intensification (Lambin et al., 2000; Armsworth et al., 2006), desertification (Pando-Moreno et al., 2004), urbanization of productive agricultural land (Seto et al., 2000; Seto et al., 2002; Seto et al., 2010) and the causes and consequences of farm abandonment (Mottet et al., 2006; Diaz et al., 2011). As a science and subject for de- bate on global sustainability issues, land use/land cover studies seek to motivate greater observation and moni- toring of land changes. It also promotes understanding of changes as a coupled human–environment system, development of spatially explicit models of land change and integrated assessments of system outcomes, such as vulnerability, resilience, or sustainability (Turner et al., 2007). Across the Arabian Peninsula little attention has been paid to the driving forces and consequences of land use/land cover change. This is remarkable given the dra- matic economic and societal changes that have taken place in these “high-income less developed” (Odedo- kun, 1996) countries in relatively short periods of time; countries where rates of population growth are globally amongst the highest. The mean rate of population in- crease for the Gulf Cooperation Council (GCC) coun- tries, plus Yemen, is variously estimated at 2.2 or 2.3% الديناميكية الزمانية واملكانية للتغريات يف استخدامات األراضي نتيجة الضغوط اخلارجية: زراعة البيوت احملمية كمثال مايكل ديدمان1* وعبداهلل السعدي1 ومالك الوردي2 وخليفة الكيومي3 ووليام ديدمان4 وفهد آل سعيد1 Abstract. Further from the northern coast of Oman new farm developments were more frequent than closer to the coast; they were also larger. The density of farms was highest close to Muscat although the distance enclosing 50% of farms had shifted away from Muscat during the study period. The dominance of Muscat is likely to be related to access to markets and infrastructure development. The increase in groundwater salinity was also highest close to Muscat and may be responsible for the shift in greenhouse density. Salinization of groundwater is severe close to the coast and was responsible for the reduced density of greenhouses near the coast. Land abandonment was highest close to Muscat and to the coast, reflecting changes in groundwater salinity and urbanization pressure. Less evidence was available for a direct shift from farmland to urban land use. Recent urban developments were largely located in areas already aban- doned by agriculture. The paper also discusses likely future trends in land use change given that Oman’s population is increasing at over 2% annually and demand for urban land is increasing. The identification of a “salinity corridor” within which much of the future land use change may occur is discussed. Keywords: Land cover; Greenhouse protected cropping; Agricultural intensification; Groundwater salinity; Urban- ization; Transport infrastructure. املســتخلص: تناقــش هــذه الورقــة التغــر احلاصــل يف توزيــع املــزارع يف حمافظــي مشــال وجنــوب الباطنــة، وبينــت الدراســة بــأن إنشــاء املــزارع اجلديــدة بعيــدا عــن الســاحل هــو األكثــر شــيوعا حاليــا ومتتــاز هــذه املــزارع أيضــا مبســاحات زراعيــة أكــر، وبالرغــم مــن أن املــزارع بــدأت بالتحــول بعيــدا عــن مســقط أوضحــت الدراســة بــأن أكثــر املــزارع مــا زالــت ترتكــز قريبــا منهــا وذلــك بســبب ســهولة الوصــول لألســواق والبــى التحتيــة. وقــد تكــون الزيــادة الشــديدة يف ملوحــة امليــاه اجلوفيــة هــي الســبب يف التغــر يف كثافــة البيــوت احملميــة بعيــدا مــن مســقط والســاحل. وأدت هــذه الزيــادة الشــديدة يف ملوحــة امليــاه اجلوفيــة والضغــط العمــراين إىل هجــر األراضــي الزراعيــة والتخلــي عنهــا واســتخدامها كمناطــق تطــور عمــراين وحضــري جديــدة. وتناقــش هــذه الورقــة أيضــا االجتاهــات املســتقبلية يف التغــر يف اســتخدام األراضــي خاصــة يف ظــل تزايــد الطلــب علــى املناطــق العمرانيــة بســبب الزيــادة الســكانية والــي تقــدر باثنــن يف املئــة ســنويا وحتديــد ممــر امللوحــة الــذي قــد حتــدث فيــه كل هــذه التغــرات يف اســتخدام األراضــي. الكلمات املفتاحية: الغطاء األرضي، الزراعة احملمية، التكثيف الزراعي، ملوحة املياه اجلوفية، التحضر، البنية التحتية للنقل 34 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping annually (CIA, 2010; UN, 2007). Omar et al. (1998) have described a scenario in Kuwait where irrigated desert lands, with minimal rainfall and loss of natural vegeta- tion to crop production activities, are prone to soil ero- sion, sand encroachment and increasing both soil and groundwater salinity. This pattern of change is reflected, at least in some of its manifestations, across the region. Salinization of irrigated land is an issue especially where coastal aquifers are depleted by agriculture and other ac- tivities at a rate that is substantially faster than rain-me- diated recharge. Seawater intrusion into the aquifer due to excessive pumping is a direct and well documented consequence not just in the Middle East (Kacimov et al., 2009) but also elsewhere where river water levels have fallen (Kotera et al., 2008). Salinization might therefore be considered as a potentially major driver behind land use change. As an increasingly important sector of land-based agriculture, greenhouse cropping has expanded rapidly since 2000. The technology was introduced into Oman in the late 1980s to help relieve some of the constraints facing the agricultural sector, in particular the hot cli- mate and shortage of water (MAF, 1994). Given the rap- idly changing dynamics of greenhouse utilization, the sector might be considered as a proxy for intensive agri- culture and thus serve as a tool for studying the vectors of land-use change, at least within the Al Batinah region where most greenhouses are located. Significant changes in the total number of green- houses have been brought about because many farm- ers have realized the importance of this technology for their financial benefit. Furthermore, the Ministry of Agriculture and Fisheries (MAF) has encouraged land use change by offering monetary incentives in the form of subsidies for greenhouse construction, resulting in a boom in development conceptually similar to that ob- served elsewhere (MacLeod and Moller, 2006; see also Lambin et al., 2001) and perhaps best expressed as ag- ricultural intervention towards productivism as defined by Wilson and Rigg (2003). Although in New Zealand subsidies seemingly caused intensification (MacLeod and Moller, 2006), in Oman there is a potential disjoin between so-called intensification represented by green- house construction and the destination of extensive ag- riculturalists displaced by greenhouse crop producers. The Al Batinah Governorates represent the most ag- riculturally active region of Oman; it is subdivided into administrative districts (Wilaya, sing. = Wilayat, Fig. 1a) each having their own local markets, but with the west- ern-most of these closer to Dubai in the United Arab Emirates than to Muscat. The marketing of farm produce from these western Wilaya is likely to be influenced by Dubai as much as by Muscat (Zekri, 2010). Across most of the region the majority of greenhouse crop produc- tion is marketed locally or at Al Mawaleh central market in Muscat Municipality, close to the high-density popu- lation of the capital area (Zekri, 2010). Al Mawaleh is a significant marketing outlet for all fruit and vegetables produced across Oman, handling in excess of 10,000 t of fruit and vegetables monthly and integrating with other, smaller markets in Muscat and elsewhere (Omezzine et al., 2002). Al Mawaleh central market, specifically, and the urban spread of Muscat in general, might therefore be expected to emerge as one of the major driving forc- es behind major land use change at the rural/peri-urban interface close to the capital area. Al Batinah is very much a transition zone where change is unlikely to be simple and unidirectional, but complex and in all probability influenced by a multiplic- ity of regional as well as local factors (Amanor and Pabi, 2007). The urbanization of agricultural land, especial- ly around the periphery of the capital area is exerting pressures on current land use distribution. Oman’s population is increasing at approximately 2% annually (CIA, 2010; UN, 2007) greatly increasing the demand for housing. Oman is committed to diversifying the econ- omy away from hydrocarbon related exports (Fasano and Iqbal, 2003) and as non-hydrocarbon revenues have grown (IMF, 2008), the recent years have seen increases in both the large and small to medium sized industrial enterprises, further increasing demand for building land beyond the common boundary shared by Muscat and Wilayat Barka. Such rapid urbanization might therefore be identified as a major driving force behind land use change in Al-Batinah. Salinity is a threat to the permanence of irrigated agri- culture in arid and semi-arid regions of the world wheth- er rainfall induced (Asseng et al., 2010) or brought about by seawater intrusion into local aquifers (Ghassemi et al., 1997). The latter inductor is dominant in Oman and in similar regions of the world. The over abstraction of groundwater for irrigation purposes has resulted in high water salinity in many parts of Oman, especially in the coastal zone of Al-Batinah, due to sea water intrusion. The Water Resources Master Plan (MWR, 2000) report- ed that the total deficit in fresh water for some parts of Al-Batinah amounted to 92.8 Mm3 and sea water intru- sion accounted for 45.8 Mm3 of water recharged into the aquifers. Al-Barwani and Helmi (2006) studied water salinity changes and rainfall in South Al-Batinah region from 1989 to 2005 and reported that salinity intrusion and groundwater salinity has increased despite the high rainfall events between 1995 and 1997. They have also reported that the area covered by high water salinity (>16 dS·m-1) has increased from 17% in 2000 to 32% in 2005 in Barka, while it has increased from 29% to 38% in As-Suwayq. The fresh water zone (< 2 dS·m-1) in both areas has decreased to 0 km2 in 2005. Given that most greenhouse growers in Oman use soil-based cultivation techniques (ICARDA, 2002) and that irrigation water with an electrical conductivity (ECw) value greater than 1.5 dS·m-1 (equivalent to 1.5 mmhos·cm-1) is considered poor quality for most greenhouse crops. Salinization clearly has the potential to be a third major driving force 35Research Article Deadman, Al-Sadi, Al-Wardi, Al-Kiyumi, Deadman, Al Said for land use change. Under investigation is the proposal that externalities have influenced greenhouse expansion projects. Specif- ically, a series of testable hypotheses are established to determine the extent, if any, of attraction or repulsion effects from (1) proximity to local markets and popula- tion centers, (2) irrigation water salinity and proximity to the coast (and thus higher groundwater salinities) (3) urban expansion and (4) transport infrastructure (Barka district only). Future agricultural zone planning for land use change depends on current knowledge of the spatial distribution of new greenhouse developments and an intuitive understanding of the sustainability of high den- sities of greenhouses in certain areas. To achieve this it is necessary at an early stage, to observe and monitor land change to gain an understanding of these changes as a coupled human–environment system, prior to the development of spatially explicit models of land change (Turner et al., 2007). As Amanor and Pabi (2007) have pointed out, land use change is unlikely to be represent- ed by a simple or linear evolution but is more probably represented by complex interactions between human and environmental influences, including spatial and temporal oscillations between extensive and intensive agriculture. In the current study only those districts through which the current coastal highway passes were includ- ed. This coastal highway links Muscat with Shinas, leads onward to Dubai and Abu Dhabi in the UAE and locally spawns numerous off-shoots, especially in those Wilaya closest to Muscat. The coastal highway represents an infrastructure interface and therefore its impact on lo- cal land use change could be significant (Lambin et al., 2000). These Wilaya also have rapidly changing ground- Table 1. Number and area of active farms and numbers of greenhouses (2001 - 2009), newly active farms (2004 - 2009) and farms abandoned (2004 - 2009) in each Wilayat (figures in brackets represent percent of total). 2001 2002 2003 2004 2009 New Abandoned Farm number Barka 30 (-49.2) 52 (-57.1) 54 (-54) 69 (-56.1) 105 (-49.5) 51 (-41.8) 18 (57. 8) Al Musanaah 5 (-8.2) 8 (-8.8) 9 (-9) 9 (-13) 23 (-10.8) 17 (-13.9) 1 (8. 9) As Suwayq 5 (-8.2) 5 (-5.5) 7 (-7) 9 (-13) 33 (-15.6) 25 (-20.5) 0 (0) Al Khaburah 7 (-11.5) 7 (-7.7) 8 (-8) 12 (-9.8) 15 (-7.1) 7 (-5.7) 5 (-11.1) Saham 6 (-9.8) 9 (-9.9) 9 (-9) 10 (-8.1) 18 (-8.5) 9 (-7.4) 1 (-4.4) Sohar 4 (-6.6) 5 (-5.5) 5 (-5) 5 (-4.1) 10 (-4.7) 6 (-4.9) 1 (-2.2) Liwa 1 (-1.6) 1 (-1.1) 1 (-1) 3 (-2.4) 3 (-1.4) 2 (-1.6) 2 (-4.4) Shinas 3 (-4.9) 4 (-4.4) 6 (-6) 5 (-4.1) 5 (-2.4) 5 (-4.1) 5 (-11.1) Total 61 91 100 123 212 122 45 Greenhouse number Barka 133 (-43.8) 340 (-58.2) 362 (-57.7) 458 (-60.6) 825 (-51.5) 402 (-50.6) 182 (-56.9) Al Musanaah 22 (-7.2) 63 (-10.8) 64 (-10.2) 64 (-8.5) 130 (-8.1) 82 (-10.3) 82 (-25.6) As Suwayq 16 (-5.3) 16 (-2.7) 19 (-3) 31 (-4.1) 292 (-18.2) 153 (-19.3) 0 (0) Al Khaburah 51 (-16.8) 49 (-8.4) 50 (-8) 56 (-7.4) 130 (-8.1) 29 (-3.6) 9 (-2.8) Saham 38 (-12.5) 58 (-9.9) 59 (-9.4) 61 (-8.1) 111 (-6.9) 46 (-5.8) 6 (-1.9) Sohar 30 (-9.9) 32 (-5.5) 27 (-4.3) 32 (-4.2) 75 (-4.7) 46 (-5.8) 4 (-1.3) Liwa 2 (-0.7) 2 (-0.3) 2 (-0.3) 14 (-1.9) 7 (-0.4) 6 (-0.8) 1 (-0.3) Shinas 12 (-3.9) 24 (-4.1) 42 (-6.7) 36 (-4.8) 31 (-1.9) 31 ((3. 9)) 36 (-11.3) Total 304 584 627 756 1601 795 320 Farm area Barka 763.8 (-62.3) 860.3 (-61.2) 871.1 (-58) 1179.1 (-56.1) 1246.5 (-33.3) 533.5 (-24.2) 466 (-82.2) Al Musanaah 43.9 (-3.6) 77.7 (-5.5) 90.3 (-6) 90.3 (-4.3) 265.2 (-7.1) 205.7 (-9.3) 30.9 (-5.4) As Suwayq 278.3 (-22.7) 278.3 (-19.8) 331.4 (-22.1) 393 (-18.7) 880.8 (-23.5) 487.8 (-22.1) 0 (0) Al Khaburah 46 (-3.8) 68.7 (-4.9) 68.7 (-4.6) 92 (-4.4) 138.8 (-3.7) 66.9 (-3) 20.1 (-3.5) Saham 52.7 (-4.3) 72.5 (-5.2) 94.2 (-6.3) 107.3 (-5.1) 162.4 (-4.3) 70 (-3.2) 14.8 (-2.6) Sohar 21.4 (-1.7) 21.4 (-1.5) 21.4 (-1.4) 203.3 (-9.7) 977.3 (-26.1) 777.5 (-35.2) 3.4 (-0.6) Liwa 6.2 (-0.5) 6.2 (-0.4) 0 (0) 10.9 (-0.5) 21.5 (-0.6) 17.3 (-0.8) 6.7 (-1.2) Shinas 13.4 (-1.1) 21.5 (-1.5) 25.2 (-1.7) 25.2 (-1.2) 49.9 (-1.3) 49.9 (-2.3) 25.2 (-4.4) Total 1225.7 1406.5 1502.2 2101 3742.4 2209 567.1 36 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping water salinity (Fig. 1b) as well as pressures of land use change through urbanization. They are also the districts within which over 95% of greenhouses in Al Batinah are located. Materials and methods In 2001, 2002, 2003, 2004 and 2009, a GPS unit was used to collect geospatial information about each farm with active greenhouses in each Wilayat. For each farm the number of productive greenhouses was recorded. During each survey note was taken of new farms and of farms that had been abandoned since the previous sur- vey. For a random selection of 100 farms in the 2009 survey, location and greenhouse number was verified against Google Earth® satellite imagery. All farm bound- aries were fixed and farm sizes determined within Goo- gle Earth®. Within the GIS environment, Wilaya bound- ary information was used to separate farms into the various administrative districts (Fig. 1a) and Al Mawaleh central market was used as a fixed reference point for distance calculations. Distances from the entrance of each farm to the nearest paved road were similarly es- timated. A B Figure 1. A. Bluemarble Next Generation image of North- ern Oman showing administrative districts (Waliya) (al- tered after NASA World Wind); inset shows study area within the Arabian Peninsula; B. Kriged model of ground- water salinity (dS·m-1) for the 30 km wide Al Batinah coastal strip with the position of the existing main high- way. Table 2. Logistic model estimates for distance (nearest km) to 50% of cumulative farm, greenhouse number and farm area from Al-Mawaleh central market. 2001 2002 2003 2004 2009 New farms Abandoned farms % Farms (R², p) 42 (.912, 0.001) 47 (.943, 0.001) 49 (.948, 0.001) 51 (.963, 0.001) 60 (.989, 0.001) 65 (.987, 0.001) 46 (.890, 0.01) % Greenhouses (R², p) 48 (.884, 0.001) 53 (.968, 0.001) 55 (.971, 0.001) 54 (.968, 0.001) 57 (.985, 0.001) 56 (.980, 0.001) 38 (.760, 0.05) % Farm area (R2, p) 42 (.927, 0.001) 44 (.945, 0.001) 47 (.946, 0.001) 54 .964, 0.001) 81 (.929, 0.001) 93 (.889, 0.01) 27 (.720, 0.05) 37Research Article Deadman, Al-Sadi, Al-Wardi, Al-Kiyumi, Deadman, Al Said Proximity to local markets, population centres and the coast The sigmoidal cumulative percent farm number, farm area and cumulative percent greenhouse number at in- creasing distances from Al Mawaleh within Muscat were calculated and modeled using a logistic function (Equ. 1) to estimate numbers of farms and greenhouses within fixed distance intervals. y = k 1−be −rx( )( ) (1) Where y is the estimated number of farms or greenhous- es; k is the upper asymptote here fixed to 100%; b and r are constants of regression and x is distance from Al- Mawaleh market. Proximity to and effects of groundwater salinity Farm number, farm area and greenhouse number with- in 1 km-wide intervals from the coast were calculated in GIS. Because greenhouse numbers were skewed to- wards the coast, cumulative sigmoidal totals were fit to the Gompertz function as an asymmetric sigmoidal model with a point of inflection at 100/e (Equ. 2). y = k e −be −rx( )( )⎛ ⎝⎜ ⎞ ⎠⎟ (2) Where y, k, b, r and x are as in equation 1. During the 2004 survey, irrigation water was collect- ed from all farms with active greenhouses and water salinity (ECw) was measured. In 2005, Ministry of Re- gional Municipalities and Water Resources’ data for 937 monitoring wells and boreholes was kriged within Arc- GIS 9.3 3-D analyst, using the ordinary kriging method with a spherical model, to provide spatial information on estimated groundwater salinity (dS·m-1, Fig. 1b). This enabled correlations to be made between kriged ground- water salinity for farm locations and salinity of collected irrigation water samples to determine likelihood values for the use of water resources by farmers alternate to groundwater supplies. The special case of Wilayat Barka Given the proximity of Wilayat Barka to Muscat and the disproportionately large number of greenhouses in that district, a detailed study was conducted to examine the interactions between farm and greenhouse location and changing salinity levels, urbanization and infrastructure development. Cumulative farm and greenhouse num- ber and farm area was related to distance from Mus- cat (Al Mawaleh) and the northern coast and to kriged groundwater salinity. In addition, areas of individually identified parcels of urban land were estimated for 2005 using Landsat ETM+ imagery and 2009 using Google Earth® images to provide GIS-based data. Cumulative urban area was modeled using the logistic function (Equ.  1). Farm and greenhouse distribution relative to the network of paved roads was quantified by buffering roads to provide distances between farm entrance and the nearest paved road. Finally, a network of 500m2 grids was used to cover the surface area of Wilayat Bar- ka and urban and farmland land use cover was estimat- ed from satellite imagery data for 2005 and 2009 (urban area) and 2004 and 2009 (farm area). Results Farm and greenhouse numbers and farm area During the study period the number of farms with green- houses in Al-Batinah increased from 61 in 2001 to 212 in 2009 and the total number of greenhouses over the same period increased from 304 in 2001 to 1601 in 2009 (Table 1). The mean number of greenhouses per farm was 4.98 in 2001, increasing to 7.53 in 2009. In 2001 al- most 30% of farms (19/61) had only a single greenhouse; in 2009 the modal greenhouse number had increased to 4.0. In 2001 only 8 farms had more than 10 greenhouses (maximum = 25), in 2009 53 farms had more than 10 greenhouses (maximum = 54). Furthermore, of the 9 farms with 30 or more greenhouses in 2009, more than half of these had commenced greenhouse crop produc- tion post 2001. Between 2004 and 2009 32 farms ceased greenhouse crop production, removing 320 greenhous- es from the productivity arena (Table 1). The highest Table 3. Gompertz model estimates for distance (nearest 0.1 km) to 50% of cumulative farm, greenhouse number and farm area from the northern coast of Oman. 2001 2002 2003 2004 2009 New farms Abandoned farms % Farms (R², p) 5.0 (.992, 0.001) 5.1 (.990, 0.001) 5.1 (.988, 0.001) 5.2 (.990, 0.001) 5.3 (.996, 0.001) 5.2 (.999, 0.001) 4.5 (.990, 0.001) % Greenhouses (R², p) 5.0 (.990, 0.001) 5.1 (.991, 0.001) 5.1 (.992, 0.001) 5.2 (.990, 0.001) 5.6 (.990, 0.001) 5.8 (.982, 0.001) 4.3 (.982, 0.001) % Farm area (R2, p) 4.7 (.984, 0.001) 4.8 (.984, 0.001) 4.8 (.985, 0.001) 4.9 (.985, 0.001) 4.5 (.992,0.001) 5.5 a (.998, 0.001) 4.1 (.982, 0.01) 5.3 a (.996, 0.001) 4.2 (.977, 0.01) a Recalculation following exclusion of two farms (see text for details) 38 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping level of abandonment was in Barka (57.8% of all aban- donments), followed by Al Khaburah and Shinas (11.1%) and then Al Musanaah (8.9%). The 2009 results showed that the majority (62%) of farms with greenhouses had land areas less than 10 ha (Fig. 2); less than 3% of farms were larger than 50 ha. Of particular relevance here is the overall decline in the preeminence of Barka in terms of farm number, declining from a high of 57% of all farms in 2002 to less than 50% of farms in 2009 (Table 1). This decline was reflected in the relatively low number (41.8%) of new-start farms in this district contained with a large proportion of total farm abandonments (57.8%). Although over 50% of new greenhouses were located in Barka, this represented less than 25% of new farm area, suggesting the development of relatively small farms with heavy emphasis on protected agriculture. Eighteen farms (representing 57.8% of the total) were abandoned in Barka, this represented 82.2% of the total agricultural land area lost between 2004 and 2009, representing rel- atively large farming areas. The situation for new farm developments in Sohar shows an opposite trend to that for Barka. In terms of land area, farms in Sohar rep- resented 26.1% of total area in 2009 and 35.2% of total land area brought into protected agriculture between 2004 and 2009. Comparatively, the number of farms and greenhouses in Sohar in 2004 and 2009 was low; a small number of large, but diversified, farms are now operating in this district. Two further observations warrant com- ment. A high proportion of farm abandonments were in Shinas (11.1% of total farm number and 11.3% of total greenhouse number), yet the contribution of these farms to total abandoned farm area was low (4.4%), suggesting the loss of small farms with high greenhouse to land area ratios. Secondly, there was a significant increase in the prominence of As Suwayq district. In 2002 As-Suwayq accounted for 5.5% of farms and 2.7% of greenhouses; by 2009 the same district accounted for 15.6% of total farm number and 18.2% of total greenhouse number. New start farms in As Suwayq represented 20.5% of total new farms and 19.3% of newly constructed greenhouses on these new farms and between 2004 and 2009 no farms were abandoned in As Suwayq (Table 1). 3.2. Proximity to Muscat Based on the fitted logistic model, the estimated distance from Muscat (Al Mawaleh market) to 50% (half distance) of cumulative number of farms, number of greenhouses and total land area for the years of the survey, togeth- er with half distances for new (between 2004 and 2009) and abandoned (post 2004) farms are shown in table 2. The half distance for all indicators extended to greater distances from Muscat during the period between 2001 and 2009 - from 42 to 60 km, from 48 to 57 km and from 42 to 81 km respectively for farm number, greenhouse number and total farm area, respectively. This appears to indicate a movement westwards (away from Muscat) of active greenhouse crop production and is reflected in the half distances for farm number, greenhouse number and farm area for new farms which were 65, 56 and 93 km respectively, whilst those for abandoned farms were Table 4. Percent farm number, greenhouse number and farm area with kriged groundwater salinities above 4.2 dS m-1, based on cumulative totals at increasing estimated salinities. 2001 2002 2003 2004 2009 New farms Abandoned farms % Farms (R², p) 60.8 (.995, 0.001) 55.8 .995, 0.001) 55.2 (.995, 0.001) 52.0 (.997, 0.001) 37.2 (.996, 0.001) 32.5 (.996, 0.001) 67.3 .990, 0.001) % Greenhouses (R², p) 45.3 (.984, 0.01) 37.6 (.984, 0.01) 37.4 (.988, 0.01) 36.9 (.987, 0.01) 22.1 (.998, 0.001) 22.9 (.998, 0.001) 60.6 (.963, 0.05) % Farm area (R2, p) 53.3 (.933, 0.05) 49.9 (.940, 0.05) 48.2 (.940, 0.05) 39.8 (.944, 0.05) 14.3 (.989, 0.01) 12.7 (.975, 0.01) 74.1 (.882, 0.01) Table 5. Logistic model estimates for distance (nearest km, within Wilayat Barka) to 50% of cumulative farm, greenhouse num- ber and farm area from Al Mawaleh central market. 2004 2009 New farms Abandoned farms % Farms (R², p) 28 (.961, 0.01) 29 (.945, 0.05) 29 (.933, 0.05) 25 (.977, 0.01) % Greenhouses (R², p) 28 (.963, 0.01) 29 (.944, 0.05) 26 (.948, 0.05) 27 (.935, 0.05) % Farm area (R2, p) 28 (.941, 0.05) 31 (.939, 0.05) 31 (.933, 0.05) 24 (.922, 0.05) 0 10 20 30 5 10 15 20 25 30 35 40 45 50 >50 Farm size (ha) Pe rc en ta ge o f t ot al fa rm s Figure 2. Frequency distribution of land area for farms with greenhouses in Al Batinah region. 39Research Article Deadman, Al-Sadi, Al-Wardi, Al-Kiyumi, Deadman, Al Said 46, 38 and 27 km. For all years and categories and for all indicators, the exponential model was an accurate and significant fit for the data (Table 2). The data in table 2 appear to corroborate that of table 1 and would fit with an increase in the relative contribution of farms in the As Suwayq district and a decrease for Wilayat Barka, es- pecially over the latter part of the survey period. 3.3. Distance from the coast and groundwater salinity Based on the 2009 data there was an increase in mean farm size with increased distance from the coast; farms adjacent to the coast had a mean area of 5.5 ha whilst those 5 km distant from the coast had a mean area of 17.4 ha. Beyond 5 km there was relatively little discern- able change in mean farm size (Fig. 3). For all years and for new and abandoned farms and for farm number, greenhouse number and farm area the fit of the Gompertz model was a significant reflection of the collected data (Table 3). Both estimated farm num- ber and greenhouse number half distances increased between 2001 and 2009 from 5.0 to 5.3 and from 5.0 to 5.6 km respectively, reflecting the southward movement of active protected agriculture production. In the case of farm area, the half distance decreased from 4.7 km in 2001 to 4.5 km in 2009, with a half distance of 4.1 km for new farms. This was primarily a consequence of two farm (ID 173 and ID 174), located 2 and 3 km from the coast, and with areas of 182 ha and 756 ha respectively commencing production in the Wilayat of Sohar, some 168 km and 170 km distant from Al Mawaleh in regions only moderately affected by salinity (Table 1). When these farms are excluded from the analysis the half dis- tance for 2009 farms increased to 5.5 km and for new farms it increased to 5.3 km (Table 3). Farms close to the coast contributed disproportionately to the number of abandoned farms (50% within 4.5 km), loss of green- houses (4.3 km) and loss of farm area (4.2 km). 0 5 10 15 20 25 2 3 4 5 6 7 8 9 10 11 12 13 Distance from the coast (km) M ea n fa rm s iz e (h a) Figure 3. Mean farm size in Al-Batinah region in relation to distance from the northern coast of Oman. y = −0.3739x + 7.1364 R 2 = 0.829, p < 0.001 5 10 15 20 4 8 12 Distance from the coast (km) E st im at ed g ro un dw at er s al in ity d S /m Figure 4. Kriged groundwater salinities for farm locations (small circles) in Al-Batinah region relative to distance from the coast. Large circles represent mean farm size for each 1 km interval. 5 10 15 20 50 100 150 200 Distance from Al−Mawaleh (km) E st im at ed g ro un dw at er s al in ity d S /m Figure 5. Kriged groundwater salinity values for farm locations in Al-Batinah region (small circles) in relation to distance from Al-Mawaleh central Market. Large circles represent mean values for 5 km intervals. 40 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping Although the relationship between kriged ground- water salinity, based on kriged 2005 borehole data, and irrigation water salinity (2004) taken directly from greenhouses was significant (p = 0.0004), irrigation wa- ter salinity was consistently lower than the estimated groundwater salinity (p(intercept) < 0.001), apparently confirming the widespread practice of importing pota- ble water to supplement well water to minimize crop toxicity damage. There is no evidence to confirm that farm wells and boreholes are accessing the same depth of water. The highest kriged groundwater salinity at farm lo- cations was in excess of 20  dS·m-1 (Fig. 4); many farms having salinity levels above 10 dS·m-1. There was a sig- nificant (R2 = .829, p < 0.001) decrease in estimated farm mean groundwater salinity with increasing distance from the coast (Fig. 4). Kriged groundwater salinity levels were also high- est close to Muscat (Fig. 5). All farms having estimat- ed salinities above 10  dS·m-1 were within 25 km of Al Mawaleh. Further peaks of salinity (above 5  dS·m-1) were observed at regular intervals along the Al Batinah coastal belt, especially at approximately 50, 125, 150 and 200 km from Al Mawaleh (Fig. 5). Groundwater salini- ty was lowest over a greater distance within As Suwayq Wilayat, approximately 90 – 120 km from Al Mawaleh. Based on kriged groundwater salinity levels calculat- ed using 2005 borehole data (Fig. 1b), and the exponen- tial model (Equ. 1), the proportion of farms with salini- ties above 4.2 dS·m-1 where a 50% reduction in cucumber yield might be expected to occur (Ayers and Westcott, 1985) was 60.8, 55.8, 55.2, 52.0 and 37.2% in 2001, 2002, 2003, 2004 and 2009, respectively. Over 67% of farms abandoned between 2004 and 2009 had groundwater salinities above 4.2  dS·m-1; less than 33% of new farms were located in such areas of high groundwater salini- ty (Table 4). In terms of total greenhouse number and total farm area, 60.6% of all abandoned greenhouses and 74.1% of abandoned land area was located in areas with salinities above the 50% yield reduction threshold (4.2 dS·m-1); less than 23% of new greenhouses and less than 13% of the area of new farms were located in such regions. The average kriged groundwater salinity of new farms developed between 2004 and 2009 was 4.08 dS·m- 1; that of farms abandoned during the same period was 7.17 dS·m-1. Although increased groundwater salinity was under- standably correlated with farm abandonment away from greenhouse production, it was also significantly relat- ed to the extent of direct investment in new protected cropping structures (Fig. 6). Farms located in regions with low estimated groundwater salinities showed sig- nificantly increased numbers of greenhouses per farm compared with those at sites where the groundwater sa- linity was higher. At 2 dS·m-1 there was a net increase of almost 7 greenhouses per farm. 3.4. The situation in Wilayat Barka The dynamics of land use change in Wilayat Barka are the most fluid: it is the district closest to Muscat and therefore most vulnerable to increasing demand for land use change to housing and industrial developments; it also has the most greenhouses (Table 1) despite having the highest groundwater salinity levels (Fig. 5), and has the most highly developed transport infrastructure (Fig. 7a). Within Barka, during the entire study period, there were 134 farms, including those abandoned during the study, occupying a total of 1729 ha and with over 1000 greenhouses. As of 2009 there were 825 greenhouses on 105 farms occupying almost 1250 ha. The majori- ty of the active farms were located to the south of the coastal highway, and up to 14 km distant from the coast (Fig. 7b, Fig. 8). The farms that had been abandoned Table 6. Gompertz model estimates for distance (nearest 0.1 km) to 50% of cumulative farm, greenhouse number and farm area from the northern coast of Oman in Wilayat Barka. 2004 2009 New farms Abandoned farms % Farms (R², p) 6.2 (.992, 0.001) 6.4 (.996, 0.001) 6.3 (.995, 0.001) 5.6 (.985, 0.001) % Greenhouses (R², p) 6.2 (.987, 0.001) 6.9 (.994, 0.001) 7.1 (.987, 0.001) 5.1 (.979, 0.01) % Farm area (R2, p) 5.7 (.988, 0.001) 6.6 (.997, 0.001) 6.3 (.996, 0.001) 4.3 (.968, 0.05) y = −0.−0.5678x + 5.9389 R 2 = 0.393, p < 0.016 −25 0 25 50 4 8 12 16 Estimated groundwater salinity (dS/m) C ha ng e in g re en ho us e nu m be r p er fa rm Figure 6. Change in number of greenhouses per farm be- tween 2004 and 2009 (small circles) relative to increasing estimated groundwater salinity. Large circles represent the mean change in greenhouse number per farm for 1 dS·m-1 intervals. 41Research Article Deadman, Al-Sadi, Al-Wardi, Al-Kiyumi, Deadman, Al Said between 2004 and 2009 were closer to the main highway and therefore closer to the northern coast of Oman, Figs. 7c, 7d, Fig. 8). Using the Wilayat Barka data, the exponential change (equation 1) in farm and greenhouse number and farm area with increasing distance from Al Mawaleh was used to determine the distance within which 50% of the to- tal of these parameters of agricultural activity occurred (Table 5). In all cases, the exponential model was a sig- nificant model of the observed change over distance and suggested that the abandonment of farms was pro- portionately higher closer to Muscat with 50% of aban- doned farms being located within 25 km of Al Mawaleh, but 50% of all 2009 farms and 50% of new farms both being located within 29 km of Al Mawaleh. The mod- el for greenhouse number showed a similar trend, with 50% of greenhouses on abandoned farms being located closer to Al Mawaleh (27 km) than 50% of greenhouses on 2009 farms (29 km) although 50% of greenhouses on new farms were actually closer to Muscat (26 km) pri- marily because of the development of two new small farms (ID 191 and ID 4 with 40 and 32 greenhouses on 7.8 and 8.2 ha) located close to Al Mawaleh (19 and 21 km respectively) but some distance from the coast (8 and 9 km, respectively). The model for farm area sug- gested the preferential abandonment of farms close to Al Mawaleh (50% within 24 km) whilst 50% of new farm area was within 31 km of Al Mawaleh. The new farms that emerged relatively close to Al Mawaleh were, how- ever, further from the northern coast. The data for Barka suggests the preferential abandon- ment of farms closer to the coast (Table 6). The thresh- Table 7. Percent farm number, greenhouse number and farm area in Barka Wilayat with estimated groundwater salinities above 4.2 dS m-1, based on cumulative totals at increasing salinities. 2004 2009 New farms Abandoned farms % Farms (R², p) 58.0 (.992, 0.001) 48.0 (.989, 0.001) 46.8 (.983, 0.001) 73.4 (.988, 0.001) % Greenhouses (R², p) 57.8 (.971, 0.01) 42.6 (.988, 0.001) 44.3 (.905, 0.01) 71.0 (.981, 0.001) % Farm area (R2, p) 63.3 (.911, 0.01) 41.5 (.979, 0.001) 40.6 (.788, 0.05) 88.7 (.987, 0.001) Figure 7. Map of Barka Wilayat showing groundwater salinity levels related to location of land abandoned to farming and recent urban developments. 42 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping old of 50% of abandoned farm number, greenhouse number and farm area was located at 5.6, 5.1 and 4.3 km respectively, closer to the coast than for 2009 farms (6.4, 6.9 and 6.6 km) and farms developed between 2004 and 2009 (6.3, 7.1 and 6.3 km, respectively). Table 7 shows the apparent importance of ground- water salinity in determining farm abandonment and location of new farm developments. Over 73% of aban- doned farms, 71% of the cumulative greenhouse num- bers and 88.7% of the cumulative total abandoned farm areas were located in regions with estimated groundwa- ter salinities in excess of 4.2  dS·m-1. New farms were predominantly located in areas with lower (<4.2 dS·m-1) groundwater salinities (only 46.8% of farms, 44.3% of greenhouses and 40.6% of total farm area was located in regions above 4.2 dS·m-1 of ECw, Table 7). The distribution of farms with active greenhouse crop production in 2009 follows closely the distribution of paved roads in Wilayat Barka (Figs 7a, 7b). This is true not only for the main highway, but also the minor roads heading south, away from the coast. The distribu- tion of abandoned farms also appeared closely related to the pattern of transport infrastructure (Fig. 7d), es- pecially those interior roads towards the eastern part of the district. In both cases, there was a significant rela- tionship between exponential cumulative farm number and distance to paved roads (R2 > 0.93, p < 0.001) with more than 50% of farm entrances being within 500m of a paved road. As of 2009, over 600 parcels of urban land were in- dividually identified in Barka as housing developments, small industrial units, retail outlets and public service buildings. Most of this urban land was north of the coastal highway, between the road and the northern coast of Oman (Fig. 7c, Fig. 9). This contrasts signifi- cantly with the distribution of agricultural land (Fig. 8). An exponential model of cumulative total urban area relative to the coast suggested that 75% of the total was within 6.1 km of the coast (R2 = 0.967, p= 0.001) and showed only a relatively small change from 2000 when 75% of the total was within 5.9 km of the coast (R2 = 0.967, p= 0.001), suggesting urban infill rather than the development of new urban sites. This infill hypothesis is also reflected by the distribution of urban land relative to Al Mawaleh. In 2000, 75% of the cumulative urban land area was within 43 km (R2 = 0.973, p= 0.001) whilst in 2009 the threshold distance was 40 km (R2 = 0.974, p= 0.001), suggesting only a slight expansion of the ur- ban area away from Muscat and this occurring with a linear trajectory, mostly between the border of Muscat and Barka town. Analysis of the grid overlying Barka Wilayat (Figs. 7a-d) clearly shows the relationship between the kriged groundwater salinity data and active (2009) and aban- doned farms (2004-2009). Although some active farms remain in high ECw areas, especially around A’Rumais village, most were in regions with much lower salinities, and almost all are south of the highway. In contrast, the majority of abandoned farms are located in areas of high groundwater salinity either in the area close to the border with Muscat to the east, or within the zone of the high salinity tongue extending south from A’Rumais. The location of urban land cover follows closely the dis- tribution of transport infrastructure, especially on the northern side of the highway and within the Barka town urban district (Fig. 7c). Figure 7c also appears to show that urban land cover largely occupies a different set of grid squares to those occupied by abandoned farms, and as such urban land cover does not represent a change in land use from agriculture, at least in terms of intensive- ly farmed agricultural land lost since 2004. Urban ex- pansion is, rather, seen as a complex mixture of infilling within regions of exceptionally high groundwater salini- ties close to the coast and ribbon development along the main highway. This notwithstanding, there remain, as of 2009, extensive areas of land close to the coast where the groundwater salinity is extremely high but where urban development has not, so far, taken place. Discussion In the current paper greenhouse (protected) cropping is used as a proxy for high intensity agriculture. That greenhouse cropping systems are intensive means of food production may appear self-evident. They normal- ly require large amounts of labour and capital per unit area of land per year and usually involve extending the growing season of crops (Jensen and Malter, 1995). The World Bank has promoted the development of green- 0 20 40 60 −7.5 −5.0 −2.5 0.0 Distance from the main highway (km) G re en ho us e nu m be r Figure 8. Number of greenhouses on farms in Barka Wilayat, Al-Batinah region, in relation to the position of the coastal highway. Positive and negative distances are north and south of the highway respectively. Large circles represent mean position relative to the coastal highway of active farms (2009, dark green) and abandoned farms (pre 2009, light grey). 43Research Article Deadman, Al-Sadi, Al-Wardi, Al-Kiyumi, Deadman, Al Said house production systems through development aid, stating “the removal of trade barriers, coupled with growing consumer demand for quality produce all year round, has further stimulated this move towards high value, intensive forms of horticultural production” (Jen- sen and Malter, 1995). The sustainability of high intensi- ty greenhouse production systems might be a less easily resolvable debate (Franze and Ciroth, 2011; Benito et al., 2009; Downward and Taylor, 2007), especially in arid and semi-arid regions where water availability for irriga- tion is often the limiting factor for agricultural develop- ment. In a life-cycle analysis of the environmental im- pact of greenhouse crop production systems, Muñoz et al. (2008) evaluated water consumption per kg of tomato as approximately 50% less for greenhouse production compared to open field production. Yet as Pandey et al. (2002) point out from a study in Niger, it is the appro- priate use of agriculture intensification technology that is vital in sustainable, increased crop production. How- ever, improvements in irrigation efficiency may merely induce growers to irrigate for longer (Peterson and Ding, 2005). Meanwhile, in the arid arena of Oman, the pro- duction intensification through greenhouse construc- tion argument rests on the need to extend the growing season in a pincer movement, starting earlier after the end of summer and pushing the cessation of production later into the spring, relative to open field production. Unless water use efficiency is improved, more irrigation will be applied per year. The results show that most farms with greenhouses are less than 10 ha (Fig. 2) and that mean farm size in- creases with distance from the coast (Fig. 3). Al Bati- nah is a region based on small farm cultivation and this has, in part been influenced by Islamic inheritance law whereby land is fragmented as it passes between gener- ations (Zekri, 2010). Older farms are, in general, nearer to the coast and have thus been more fragmented over a longer period of time than more recently established farms inland. Alternatively, as Zekri (2010) suggests, other socio-economic causes of fragmentation exist, in- cluding the selling of parts of farms for urban develop- ment. The danger here is that continued fragmentation reduces agricultural sustainability (Fan and Chan-Kang, 2005). When modeling the expansion of agriculture in Ken- ya’s Eastern Arc Mountains, Maeda et al. (2010) iden- tified the main factors driving the spatial distribution of land brought into agriculture as distance to markets, proximity to already established agricultural areas and distance to roads. The dominance of the Muscat capital area in determining the distribution of intensive agri- cultural production units is similarly apparent here. Al- though there was an outward extension, between 2004 and 2009, of the distance within which 50% of cumulative farms (42 km to 60 km), greenhouses (48 km to 57 km) and farm area (42 km to 81 km), and the 50% threshold for new farms developed between 2004 and 2009 was further from Muscat, the overwhelming majority of in- tensive agriculture remained close to the main market in Al Mawaleh (Table 2). Transport infrastructure was also a clear driver for land use: 50% of farms with greenhous- es were within 500 m of the nearest paved road (Fig. 7b). Notwithstanding the market and infrastructure driv- ers of land use for intensive (greenhouse) agricultural production, clustering of land use components was clear (see Fig. 7b for Wilayat Barka). Land use clustering is a well-documented social–spatial externality (Lewis et al., 2008) resulting from processes related to changes in information flows (diffusion of information, Foster and Rosenzweig, 1995) or imitation (Schmit and Rounsevell, 2006), transaction costs, fixed costs, infrastructure, and other factors. But clustering was also observed in aban- donments away from intensive agriculture (Fig. 7d) and abandonments were disproportionately closer to Mus- cat; in this case the principal drivers are more likely to be biophysical, mediated through socio-economic, rather than purely socio-economic. In Oman, there is little direct evidence of a linear mo- mentum from agricultural land use to urban land use. Even in Barka, where urban pressures are greatest, Fig- ure 7c shows that co-habitation by abandoned farms and urban land parcels within the individual 250,000 m2 (25 ha) cells superimposed on the Wilayat, is low. In Barka, the growth of urban developments has almost entirely been restricted to the strip of land between the existing highway and the coast (Figs. 1b, 7, 8, 9). This land area is mostly dominated by highly saline soils previously aban- doned by productivist agriculture (Fig. 7d). As intensive agriculture has been driven out of what might be called a “saline corridor”, it has been replaced by a mixture of land cover types, including urban developments and abandoned agricultural lands reclaimed by natural veg- etation and quasi pre-productivist rural land use with extensively grazed goat herds dominating. In Southern Chile Diaz et al. (2011) recorded an almost 50% level of agricultural land abandonment between 1985 and 2007 0.0 0.3 0.6 0.9 1.2 −10 −5 0 5 Distance from the main highway (km) A re a (h a) Figure 9. Distribution of urban land in Barka Wilayat, Al-Batinah region in relation to the position of the coast- al highway. Positive distances and negative distances are north and south of the highway respectively. 44 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping in their study area, with important drivers of the return to arboreous shrubland being identified as, amongst others, soil quality. They also identified policy-driven subsidies as an important socio-economic factor behind land abandonment. It is clear that in the Al Batinah “sa- line corridor”, soil salinity is the major driver of land use change, encouraging a trajectory towards abandonment; it is also possible that the continuing government subsi- dies offered to farmers in Oman for the construction of greenhouses are similarly driving the abandonment of land in marginal areas: growers are unlikely to expand the number of productive units when yields are at risk from high groundwater salinities unless expensive po- table water is imported. In any case, in this area small farm size may also preclude expansion. Recent literature has concentrated on land-use / land -cover as drivers of change in groundwater quality. Perhaps predictably, Singh et al. (2011) found that land use change towards urbanization and industrialization in Punjab, India resulted in a reduction in groundwater quality over 1989-2006. A similar outcome has been re- ported in Western Turkey (Sanli et al., 2009) although Twarakavi and Kaluarachchi (2006) reported an increase in groundwater quality when urbanization displaced agriculture as the dominant land use and levels of ni- trate and other agrochemical residues declined. Xu et al. (2007) examined the impact of land use on changes in groundwater nitrate quantity using spatially integrat- ed data during the 1960s to 1990s. Not surprisingly ni- trate levels rose as agriculture intensified and later fell as groundwater levels themselves dropped or land became urbanized. More interestingly vector differences in groundwater quality changes were observed for different land uses and land use changes (desert to agriculture, desert to urban and agriculture to urban). In Oman, the impact of intensive agriculture on groundwater quality close to the coast is already well documented, greater interest now lies in the consequent land use change tra- jectories. The government has adopted the expedient socio-economic position of responding to the decline in agriculture caused by the change in groundwater qual- ity by introducing an industrialization or urbanization policy along the “salinity corridor”, including the com- pulsory purchase of farm land. Whether this would have occurred in the absence of the groundwater quality change reducing the tenacity of intensive agriculture is debatable. What is clear is that the next decade is likely to bring urbanization and financial investment in infra- structure to Oman at a previously unseen scale. Major transport projects are planned for the “saline corridor”; these are certain to result in significant industrial and urban land use changes as an additional highway runs the length of the coast between the Wilaya of Barka and Shinas. Inland, a new highway and rail system is to be developed that is likely to attract urban development along the roads connecting these highways (Anony- mous, 2010) The results of the current study suggest an over- all movement of the intensive agriculture (proxied by greenhouse developments) zone inland and away from Muscat (see Fig. 7b for Barka Wilayat vectors), away from the coast with its high soil salinity and following the improving internal transport infrastructure. The largest net change in recent intensive agricultural activi- ty has been the increase, especially in terms of farm area, in Wilayat As Suwayq (Table 1). This district is the low- est in the apparent extent of groundwater salinity (Fig. 1b) and has the lowest percent area affected by salinities above 9 dSm-1 (Zekri, 2010). This momentum will need to be addressed at the socio-economic policy level to prevent a future decline of groundwater salinity in this district. Conclusion Oman is a rapidly changing country. A high rate of pop- ulation growth and policy-driven industrial diversifica- tion away from petrochemicals is encouraging urbaniza- tion at the expense of traditional agriculture. Intensive agriculture appears to be responding by moving away from areas of most rapid urban change and away from areas of highest groundwater salinity: new, large farms are being developed inland as small, traditional farms close to the coast and close to Muscat are abandoned. This movement appears to be facilitated by expanding networks of transport infrastructure. For the first time in Oman the dynamics of the changing land use have been analyzed providing researchers and others with da- tabase information to explore other vectors for change in the agricultural sector. References Al-Barwani, A. and T. Helmi. 2006. Seawater intrusion in a coastal aquifer: A case study for the area between Seeb and Suwaiq, Sultanate of Oman. Sultan Qaboos University Journal for Agricultural and Marine Sci- ences 11:55-69. Amanor, K.S. and O. Pabi. 2007. Space, time, rheto- ric and agricultural change in the transition zone of Ghana. Human Ecology 35: 51-67. Anonymous. 2010. Comprehenisive Master Plan for Al Batinah Coastal Area. Phase-1 Final Report: Analysis and Assessment of Study Area. Supreme Committee for Town Planning, Sultanate of Oman. Armsworth, P.R., G.C. Daily, P. Kareiva, and J.N. Sanchirico. 2006. Land market feedbacks can under- mine biodiversity conservation. Proceedings of the National Academy of Sciences of the United States of America 103: 5403-5408. Asner, G.P., D.E. Knapp, E.N. Broadbent, P.J.C. Oliveira, M. Keller, and J.N. Silva. 2005. Selective logging in the Brazilian Amazon. Science 310: 480-482. 45Research Article Deadman, Al-Sadi, Al-Wardi, Al-Kiyumi, Deadman, Al Said Asseng. S., A. Dray, P. Perez, and X. Su. 2010. Rainfall– human–spatial interactions in a salinity-prone agri- cultural region of the Western Australian wheat-belt. Ecological Modelling 221: 812-824. Ayers, R.S. and D.W. Westcott. 1985. Water Quality for Agriculture. Irrigation and Drainage Paper 29. Rome, FAO. 174pp. Benito, B.M., M.M. Martinez-Ortega, L.M. Munoz, J. Lorite, and J. Penas. 2009. Assessing extinction-risk of endangered plants using species distribution mod- els: a case study of habitat depletion caused by the spread of greenhouses. Biodiversity and Conserva- tion 18: 2509-2520. CIA, 2010. The World Factbook. Central Intelligence Agency Office of Public Affairs. USA, Washington. Diaz, G.I., L. Nahuelhual, C. Echeverria, and S. Marin. 2011. Drivers of land abandonment in Southern Chile and implications for landscape planning. Landscape and Urban Planning 99: 207-217. Downward, S.R. and R. Taylor. 2007. An assessment of Spain’s Programa AGUA and its implications for sustainable water management in the province of Almeria, southeast Spain. Journal of Environmental Management 82: 277-289. Fan, S.G. and C. Chan-Kang. 2005. Is small beautiful? Farm size, productivity, and poverty in Asian agricul- ture. Agricultural Economics 32: 135-146. Fasano, U. and Z. Iqbal. 2003. GCC Countries: From Oil Dependence to Diversification, Washington: Interna- tional Monetary Fund. Foster, A. and M. Rosenzweig. 1995. Learning by doing and learning from others: Human capital and techni- cal change in agriculture. Journal of Political Econo- my 103: 1176-1209. Franze, J. and A. Ciroth. 2011. A comparison of cut ros- es from Ecuador and the Netherlands. International Journal of Life Cycle Assessment 16: 366-379. Ghassemi, F., A. Close, and J.R. Kellett. 1997. Numer- ical models for the management of land and water resources salinisation. Mathematics and Computers in Simulation 43: 323-329. ICARDA. 2002. Integrated Management of Cucumber and Tomato Pests under Protected Cultivation Sys- tems. Aleppo, International Center for Agricultural Research in the Dry Areas. IMF. 2008. IMF Executive Board Concludes 2007 Ar- ticle IV Consultation with Oman. International Monetary Fund Public Information Notice (PIN) No. 08/50, April 30, 2008. Jensen, M.H. and A.J. Malter. 1995. Protected Agricul- ture: A Global Review. World Bank Technical Paper Number 253, World Bank, Washington, USA, 176 pp. Kacimov, A.R., M.M. Sherif, J.S. Perret, and A. Al-Mu- shikhi. 2009. Control of sea-water intrusion by salt-water pumping: Coast of Oman. Hydrogeology Journal 17: 541-558. Kotera, A., T. Sakamoto, D.K. Nguyen, and M. Yokoza- wa. 2008. Regional consequences of seawater in- trusion on rice productivity and land use in coastal area of the Mekong river delta. Japan Agricultural Research Quarterly 42: 267-274. Lambin, E.F., M.D.A. Rounsevell, and H.J. Geist. 2000. Are agricultural land use models able to predict changes in land use intensity? Agriculture, Ecosys- tems and Environment 82: 321-331. Lambin, E.F., B.L. Turner, H.J. Geist, S.B. Agbola, A. An- gelsen, J.W. Bruce, O.T. Coomes, R. Dirzo, G. Fischer, C. Folke, P.S. George, K. Homewood, J. Imbernon, R. Leemans, X.B. Li, E.F. Moran, M. Mortimore, P.S. Ramakrishnan, J.F Richards, H. Skanes, W. Steffen, G.D.Stone, U. Svedin, T.A. Veldkamp, C. Vogel, and J.C. Xu. 2001. The causes of land-use and land-cover change: moving beyond the myths. Global Environ- mental Change - Human and Policy Dimensions 11: 261-269. Lewis, D.J., B.L. Barham, and K.S. Zimmerer. 2008. Spatial externalities in agriculture: Empirical analy- sis, statistical identification, and policy implications. World Development 36: 1813-1829. Maeda, E.E., B.J.F. Clark, P. Pellikka, and M. Siljander. 2010. Modelling agricultural expansion in Kenya’s Eastern Arc Mountains biodiversity hotspot. Agri- cultural Systems 103: 609-620. MacLeod, C.J. and H. Moller. 2006. Intensification and diversification of New Zealand agriculture since 1960: An evaluation of current indicators of land use change. Agriculture Ecosystems and Environment 115: 201-218. MAF. 1994. Planting Cucumber and Tomato in the Greenhouses. Muscat, Ministry of Agriculture and Fisheries. Mottet, A., S. Ladet, N. Coque, and A. Gibon. 2006. Ag- ricultural land-use change and its drivers in moun- tain landscapes: A case study in the Pyrenees. Agri- cultural Ecosystems and Environment 114:296-310. Muñoz, P., A. Antón, M. Nuñez, A. Paranjpe, J. Ariño, X. Castell, J.I. Montero, and J. Rieradevall. 2008. Com- paring the environmental impacts of greenhouse versus open-field tomato production in the Mediter- ranean region. Acta Horticulturae 801: 1591-1596. MWR. 2000. National Water Resources Master Plan. Internal Report, October 2000. Ministry of Water Resources, Sultanate of Oman. 120pp. Nepstad, D.C., A. Verissimo, A. Alencar, C. Nobre, E. Lima, P. Lefebvre, P. Schlesinger, C. Potter, P. Moutin- ho, E. Mendoza, M. Cochrane, and V. Brooks. 1999. Large-scale impoverishment of Amazonian forests by 46 SQU Journal of Agricultural and Marine Sciences, 2016, Volume 21, Issue 1 Dynamics of land use changes in response to external pressures in Oman: Greenhouse cropping logging and fire. Nature 398:505-508. Odedokun, M.O. 1996. Alternative econometric ap- proaches for analysing the role of the financial sec- tor in economic growth: Time-series evidence from LDCs. Journal of Development Economics 50: 119- 146. Omar, S.A.S., T. Madouh, I. El-Bagouri, Z. Al-Mussalem, and H. Al-Telaihi. 1998. Land degradation factors in arid irrigated areas: The case of Wafra in Kuwait. Land Degradation and Development 9: 283-294. Omezzine, A., O. Al-Jabri, and H. Boughanmi. 2002. Analysis of fruit and vegetable price integration be- tween Al Mawaleh market and Dubai wholesale market. Agricultural and Fisheries Research Bulle- tin (Ministry of Agriculture and Fisheries, Muscat, Oman) 2: 5-10. Pando-Moreno, M., E. Jurado, M. Manzano, and E. Es- trada. 2004. The influence of land use on desertifica- tion processes. Journal of Range Management 57: 320-324. Pandy, R.K., T.W. Crawford, and J.W., Maranville. 2002. Agriculture intensification and ecologically sustain- able land use in Niger: A case study of evolution of intensive systems with supplementary irrigation. Journal of Sustainable Agriculture 20: 33-55. Peterson, J.M. and Y. Ding. 2005. Economic adjustments to groundwater depletion in the high plains: Do wa- ter-saving irrigation systems save water? American Journal of Agricultural Economics 87: 147-159. Sanli, F.B., Y. Kurucu, and M.T. Esetlili. 2009. Determin- ing land use changes by radar-optic fused images and monitoring its environmental impacts in Edremit re- gion of western Turkey. Environmental Monitoring and Assessment 151: 45-58. Schmit, C. and M.D.A. Rounsevell. 2006. Are agricul- tural land use patterns influenced by farmer imita- tion? Agriculture, Ecosystems and Environment 115: 113-127. Seto, K.C., R.K. Kaufmann, and C.E. Woodcock. 2000. Landsat reveals China’s farmland reserves, but they’re vanishing fast. Nature 406: 121. Seto, K.C., C.E. Woodcock, C. Song, X. Huang, J. Lu, and R.K. Kaufmann. 2002. Monitoring land-use change in the Pearl River Delta using Landsat TM. Interna- tional Journal of Remote Sensing 23: 1985-2004. Seto, K.C., R. Sánchez-Rodríguez, and M. Fragkias. 2010. The new geography of contemporary urbaniza- tion and the environment. Annual Review of Envi- ronment and Resources 35: 167-94. Singh, C.K., S. Shashtri, S. Mukherjee, R. Kumari, R. Av- atar, A. Singh, and R.P. Singh. 2011. Application of GWQI to assess effect of land use change on ground- water quality in Lower Shiwaliks of Punjab: Remote sensing and GIS based approach. Water Resources Management 25: 1881-1898. Turner, B.L., E.F. Lambin, and A. Reenberg. 2007. The emergence of land change science for global environ- mental change and sustainability. Proceedings of the National Academy of Sciences of the United States of America 104: 20666-20671. Twarakavi, N.K.C. and J.J. Kaluarachchi. 2006. Sustain- ability of groundwater quality considering land use changes and public health risks. Journal of Environ- mental Management 81: 405-419. UN. 2007. World Population Prospects: The 2006 Revi- sion, Highlights, Working Paper No. ESA/P/WP.202. United Nations, Department of Economic and Social Affairs, Population Division. Wilson, G.A. and J. Rigg. 2003. Post-productivist agri- cultural regimes and the south: discordant concepts? Progress in Human Geography 27: 681-707. Xu, Y., L.A. Baker, and P.C. Johnson. 2007. Trends in ground water nitrate contamination in the Phoenix, Arizona Region. Ground Water Monitoring and Re- mediation 27: 49-56. Zekri, S. 2010. Agriculture. In: Comprehensive Master Plan for Al Batinah Coastal Area. Supreme Commit- tee for Town Planning, Sultanate of Oman.