Abstract
The mobility of coastal dunes is characterised by bio-geomorphological responses related to change in boundary conditions, particularly sediment supply, wind and vegetation cover, as well as human activities. There remains uncertainty regarding the relative importance of these drivers on dune mobility at a global scale. In this study, trends and dominant drivers of coastal dune mobility are synthesised through the literature review focusing on shifts in dune mobility over the last century (1870–2018). In total, 176 individual dunes, with 55 dunes from the Europe-Mediterranean area, 23 from Africa, 30 from North America, 23 from South America, 20 from Oceania and 23 from Asia, are reviewed in this work. The results show that there is a worldwide trend of dune stabilisation, with 93% (164 out of 176) of the reviewed sites showing a loss of bare sand area due to an increase in vegetation cover and urbanisation expansion. Multiple factors have contributed to the stabilisation process, including (a) land-use change such as the change of traditional farming practises, coastal urbanisation and tourism development; (b) dune stabilisation projects; (c) sediment decline caused by the riverine and coastal constructions; and (d) change in climate (i.e. the decrease in windiness, and the increase in temperature and rainfall) and storms. Our results suggest human intervention played a dominant role in altering dune mobility for most dunes during the past century, while climate and storms are also important drivers, especially for dune sites with limited human activities.
I Introduction
Dune systems are important landforms occupying a large part of the world’s coastline. The evolution of coastal dunes occurs over a wide range of spatial and temporal scales (Figure 1): from the instantaneous transport of individual sand grains (Bauer et al., 2009; Jackson and Cooper, 1999), to event-driven changes such as dune erosion by storms (Claudino-Sales et al., 2008; Davidson-Arnott et al., 2012), foredune development over years to decades (Levin et al., 2017; Ollerhead et al., 2013) through to coastal plain evolution on centennial to millennial scales (Clemmensen et al., 2009; Oliver and Woodroffe, 2016; Oliver et al., 2018; Sherman and Bauer, 1993). All these scale-based dynamics are moderated by the system boundary conditions.

Spatial and temporal scales involved in coast morphological evolution (modified after Cowell and Thom, 1994). On the instantaneous scale, morphological evolution is forced by primary agents, such as wind; on the event scale, a single event such as a storm is the main force; on the engineering/historical scale, geomorphology involves composite evolution over many fluctuations of boundary conditions; on the geological scale, morphology is a response to mean trends in environmental conditions.
Sediment supply, wind and vegetation are the three principle boundary conditions for dune system formation and development (Aagaard et al., 2007; Costas et al., 2012, 2016; Delgado-Fernandez and Davidson-Arnott, 2011; Hesp, 2013; Psuty and Silveira, 2010; Pye, 1983). Other factors like storms and sea-level changes also impact dune mobility through shifts in sediment supply and vegetation (Feagin et al., 2005; Miller et al., 2010). Fluctuations in boundary conditions drive changes in dune mobility through time. On a decadal scale, the net result is that dune mobility spans a continuum: from completely bare sand and, therefore, highly mobile, to entirely vegetated and stable surfaces (Hesp and Thom, 1990; Levin et al., 2008; Tsoar, 2005).
Over the past century, human activities have played an increasingly important role in altering dune mobility (Nordstrom, 1994). For example, afforestation programmes have accelerated in dune stabilisation in several countries (Avis, 1989; Reckendorf et al., 1985), while overgrazing and trampling have destroyed vegetated surfaces, resulting in dune mobilisation (Arens et al., 2007; Delgado-Fernandez et al., 2019a; Hesp et al., 2010; Konlechner et al., 2014, 2015). Urbanisation and coastal management have also altered dune mobility through land-use change (Hernández-Cordero et al., 2018; Malavasi et al., 2013; Marcomini et al., 2017; Sytnik and Stecchi, 2015).
Alternatively, several studies have focused on the role of climate as a principal driver of increases in dune stability (Da Silva and Hesp, 2013; Jackson and Cooper, 2011). For example, a recent study shows a global ‘greening’ (increased vegetation cover) of coastal dunes since 1984 (Jackson et al., 2019). This stabilisation was attributed to the changes in climate and atmospheric composition (N, P, CO2); however, this study only considered dune sites with limited or no human intervention. Globally, many dunes are situated in coastal environments where human impact has occurred over centuries. Human interventions on the coast have intensified over the past century, and, moreover, there is no indication that these activities will cease (Nordstrom, 1994).
Dune mobility cannot be directly related to a perturbation in a single boundary condition because dune activity is an integrated biogeomorphological response to all boundary conditions and human disturbance (Delgado-Fernandez et al., 2019a). In many instances there is a lag between an event and the subsequent landform response (Houser et al., 2015; Pickart, 2013); and geomorphological responses themselves are nonlinear to boundary condition changes (Cowell and Thom, 1994; Tsoar, 2005; Yizhaq et al., 2007). This complexity has meant it has not been possible to quantitively determine the impact of potential drivers on dune mobility in a short-term time. Therefore, this review sets out to explore the dynamics of dunes either with considerable human activities or rare/limited human interventions across the globe on a decadal scale to identify: (a) whether there are global trends of dune mobility during the past century and (b) the dominant factors that have driven the contemporary landform stability.
II Methods
This study utilises existing empirical studies to synthesise the extent and causes of global coastal dune mobility. Relevant studies were identified through Web of Science (Core Collection database) using the search parameters ‘TI=((coast* dune$ OR coast* sand$ dune$ OR coast* dunefield$ OR coast* dune field$) NOT (desert dune$ OR Mars)) AND TS=(“*stabili*” OR “*moil*” OR “*activat*” OR “fixation” OR “*vegetat*” OR rainfall OR storm$ OR climat* OR wind$)) AND LANGUAGE: (English) AND DOCUMENT TYPES: (Article) AND Timespan=All years (1900-2020)’. All the returned papers were reviewed and those that did not explicitly address the change in dune mobility were omitted. Additional articles were collected by snowball sampling (by tracking the references at the back of relevant papers).
In the reviewed articles, changes in dune mobility were mainly identified by the comparison of the earlier and recent dune landforms based on aerial photos, satellite images, field surveys, postcards and photos. Information from other sources (i.e. the dune stabilisation handbook) was also used to describe the mobility change for some cases in the articles.
For this study, three dune mobility trends were defined. Specifically, dune stabilisation was defined where there is a loss in bare sand area and/or an increase in vegetation regardless of the method used to identify these changes; dune mobilisation was defined where there is a gain in bare sand area and/or a loss in vegetation; otherwise, we labelled the dune as little change in dune mobility. For dunes with different patterns in mobility change through time, only net change from the earliest year to the latest year was considered in this work. In addition, dune sites were divided into two categories. If human activities (including farming, grazing, recreation, urbanisation, revegetation, revegetation) took place on the dune, we marked this site as a human-impacted dune with considerable human interventions; otherwise, the site was labelled as a natural dune with rare/limited human intervention. Location, study period, mobility trend and the considered drivers for each dune site (if available) were collected from the literature and compared with their mobility trends.
III Global trends of coastal dune mobility
The literature review returned information on 176 individual dune systems, with 55 dunes located on the Europe-Mediterranean coast, 23 on the African coast, 30 on North America, 23 from South America, 20 from Oceania and 23 from Asia (Figure 2; Table 1). Among these dunes, most sites (82% or 144 out of 176) were affected by human activities, and 32 dune sites were considered to have rare or limited human intervention (Table 1). The period over which dunes were studied in the reviewed articles covered a wide range, from 1870 to 2018, with most studies focusing on shifts in dune mobility between 1940 and 2000 (Figure 3). Compared with the span from 1984 to 2017 in the study by Jackson et al. (2019), the average span of each study was 49 years in this work, thus allowing for multi-decadal shifts in dune mobility to be evaluated.

Global distribution of reviewed dune sites from the literature. (Black dots and red symbols present dune sites with stabilising and mobilising trends, respectively; and blue triangles show the dunes with little change in dune mobility.) For interpretation of the references to colours in this figure legend, refer to the online version of this article.
Numbers of reviewed dune sites from different regions.

Time spans of coastal dunes studied by reviewed articles in different regions. (The horizontal bars show the studied time span of dunes; colours refer to different trends of dune mobility.) For interpretation of the references to colours in this figure legend, refer to the online version of this article.
Overall, there is an increase in dune stability on a global scale during the past century (Figure 2), consistent with the results of Jackson et al. (2019). Specifically, 164 dunes (or 93% of the reviewed dune sites) showed a stabilisation trend, 10 dunes showed little change in dune mobility and two dunes showed a mobilisation trend (Table 1). More details of mobility changes and specific drivers are reviewed by different geographical regions in the following section.
IV Regional patterns of dune mobility
1 Europe-Mediterranean
Coastal dunes fringe large sections of the shorelines of Europe, with a total area of over 5300 km2 (Doody, 2008). These dunes experienced a widespread trend of stabilisation during the last century (Figure S1, supplemental material). For example, most of the active dunes in Iceland have increased in stability since 1907 as a result of planting and prohibition of grazing (Runólfsson, 1987). In the UK, previous mobile dunes caused by agricultural practises have shifted toward increased stability since the 1940s (Bailey and Bristow, 2004). For example, dunes in Wales showed a decline in bare sand area by an average of 82% between the 1940s and 2009, ranging from 41% at Gronant Dunes and Talacre Warren to 97% at Kenfig Burrows (Pye et al., 2014). Mobile dune area at Portstewart in Northern Ireland reduced by 86–96% during 1964–2004, and similar reductions by 51–84% were also observed at another 10 dune sites in the north and west Ireland in 1995–2005 (Jackson and Cooper, 2011).
In Denmark, about half the dunes on the west coast had shifted from a mobile to a forested state between the 1850s and 1930s (Clemmensen and Murray, 2006; Clemmensen et al., 2014). The large dune Råbjerg Mile on Skagen Odde was the only dune remaining active in Denmark, which also showed an increase in vegetation cover from 29% in 1887 to 94% (Anthonsen et al., 1996). Similarly, most Polish dunes have been stabilised by marram planting and afforestation since the late 19th century (Łabuz et al., 2018), and the only remaining mobile dune (the Łeba Sandbar dune in Figure S1) also showed a 54% reduction in the area of bare sand during 1932–2006 (Peyrat et al., 2009). In the Netherlands, a large extent of mobile dunes have been stabilised by forests and marram since the 1900s (Doody, 2008; Jungerius and Van der Meulen, 1988; Van Dijk, 1992). In Belgium, 37% of the mobile dunes have been lost to urbanisation since the beginning of the 20th century, whilst the rest were almost completely vegetated by the 1980s (De Raeve, 1989; Provoost and Van Landuyt, 2001).
Aquitaine in southwestern France has the largest dune system in Europe. Today, the dunes there are almost completely stabilised due to the extensive planting of marine pines in the first half of the 19th century (Clarke et al., 2002; Paskoff, 2001). The largest mobile dune in Spain (the Doñana dune system in Figure S1) also showed a decrease in dune mobility during 1956–1999 (Doody, 2008; Muñoz-Reinoso, 2018; Ojeda et al., 2005). Similar stabilisations are found in the Cíes Islands (Costas et al., 2006), Traba (González-Villanueva et al., 2013), along the Catalan coastline (Garcia-Lozano and Pintó, 2018) and on the Canary Islands (Hernández-Cordero et al., 2018). However, two transgressive dunes in southern Spain, namely the Bolonia dune and the Valdevaqueros dune (Figure S1), have shown increased mobility in recent decades (1974–2011) due to the unlimited sediment availability from the beach (Navarro-Pons et al., 2016).
Dune stabilisation also occurred along the Mediterranean coast. Dunes in the northeast and central Italy showed a 75–80% loss to urban expansion and tourism development since the 1950s (Miccadei et al., 2011; Sytnik and Stecchi, 2015). In Turkey, the Kapiköy dune in the south was highly active prior to the 1970s, but it is now stabilised by exotic forest (Akça et al., 2010). Other cases with a reduction in bare sand area were reported in Kemer and Tuzla (Çakan et al., 2011; Kemal Sönmez et al., 2009). Increase of dune stability in Israel (i.e. dunes in Ashdod-Nizanim) started since the 1940s, attributed to the land-use change combined with a decrease in windiness (Kutiel et al., 2004; Levin and Ben-Dor, 2004). Dunes in Egypt, which were active in 1955, had been removed or replaced by summer resorts and other infrastructure by 2003 (El Banna, 2004). Dunes in north Libya also showed a stabilisation trend from the early 19th century to 1984, followed by deforestation during 1984–2016 (Alsoul, 2016).
2 Africa
Dunes occur along more than 80% of the shoreline in South Africa, and they were largely mobile around the mid-18th century (Harris et al., 2011; Tinley, 1985). These dunes had increased in stability since the 1830s, following the commencement of a systematic dune stabilisation program (Avis, 1989). The earliest organised dune stabilisation efforts began in 1845 in the Western Cape with the sowing of seed and planting of seedlings directly in the sand, followed by afforestation combined with the spreading of city refuse and the large planting of marram grass (Avis, 1989, 1995). By the end of the 20th century, about 40% (1300 km out of 3110 km) of the South African coast had been stabilised (Figure S2, supplemental material) (Hertling and Lubke, 1999; Tinley, 1985). In Mozambique, dunes along the coast of Gaza province were also well-vegetated and fixed with human constructions for some locations (Miguel and Castro, 2018). Increases in dune mobility were also observed in Essaouira (Morocco), Mudug (Somalia), Port Elizabeth and Arniston (South Africa), and change in climate and atmospheric nutrients was cited as the main driver (Jackson et al., 2019). A few dunes, including the Bazaruto dune and Maputo dune in Mozambique and the Alexandria dune in South Africa, are still largely active today (Illenberger and Rust, 1988; Jackson et al., 2019; Miguel and Castro, 2018).
3 North America
Large dune systems are widely distributed along the west coast (Oregon, Washington and Baja California) of the US (Figure S3, supplemental material), which have been largely stabilised since the turn of the 20th century (Carlson et al., 1991; Cooper, 1958; Reckendorf et al., 1985, 1987). For example, mobile dune in Warrenton was a concern in the early 1930s; however, it was fixed in 1940s by a stabilisation project via sand fencing and grass planting, followed by permanent stabilisation with herbaceous or woody vegetation and other land-use restrictions (Reckendorf et al., 1985). Since then, more dunes along the west coast were subsequently stabilised utilising similar methods, like the Warrenton project (Figure S3) (Reckendorf et al., 1987). A recent study, however, has demonstrated that the stabilisation of Lanphere-Ma-le’l transgressive dune system in northern California in 1948–2016 was coincident with a shift of the Pacific Decadal Oscillation from the cool to warm phase, suggesting climate has at least partly driven dune system changes here (Pickart and Hesp, 2019).
Jockey’s Ridge in North Carolina, which is the largest dune field along the east coast of the US, also showed an increase in vegetation cover and dune stability in 1932–2003 (Mitasova et al., 2005). Increase in dune stability with a reduced migrating rate was observed on dunes in Napeague and Cape Cod during the 1930s–2000s, although parts of these dunes remain active and migrating (Abhar et al., 2015; Forman et al., 2008; and Girardi Davis, 2010). Indiana Dunas on the southern shore of Lake Michigan also showed a decrease in bare sand area from 1938 to 2008, although beach erosion and accretion alternated over different periods (Kilibarda and Shillinglaw, 2015). A loss of bare sand to the shrub encroachment caused by increased temperature was also observed on Hog Island, VA, during 1984–2016 (Huang et al., 2018).
In Mexico, coastal dunes in Playa Chachalacas (Figure S3) in the central region of the Gulf of Mexico showed an increasing trend toward stability. This was promoted by a warm and wet climate, as well as impacted by human activities (breakwater construction and urbanisation) (Martínez et al., 2019).
In southwest Canada, transgressive dunes at Wickaninnish and Schooner Cove had experienced reduction in active sand surface by 28% and 30%, respectively, due to changes in management practices and a warmer and wetter climate over 34 years (Heathfield and Walker, 2011; Walker et al., 2013). Further south, dunes in the Pacific Rim National Park Reserve have been stabilised by marram grass coupled with similar shifts in climate (Darke et al., 2013). On the east coast, the Sable Island dunes were mobile and almost free of vegetation prior to the 1950s (Cameron, 1965); however, dune mobility has decreased by 50% following an increase in vegetation cover (Tissier et al., 2013). The Greenwich Dunes on Prince Edward Island also showed increased stability over the period 1939–2005 (Mathew et al., 2010).
4 South America
Transgressive dunes are widely distributed along the Brazil coast (Figure S4, supplemental material). Most dunes along the southeast Brazil coast have increased in stability over the past century. For example, the coast of Santa Catarina State and Rio Grande do Sul State in southeastern Brazil were almost free of vegetation before the 1950s, but all of these dunes showed an increase in stability as vegetation coverage expanded (Da Silva and Hesp, 2013; Jackson et al., 2019; Martinho et al., 2010; Mendes and Giannini, 2015; Pinto et al., 2015; Seeliger et al., 2000). The large dunefield in Pequenos Lençois Maranhenses, northeast Brazil, has also shown a stabilising trend over the past 30 years (Jackson et al., 2019); Jenipabu’s Environmental Protection Area (Pinto and Fernandes, 2011) was similar. In contrast, some mobile dunes can be found in northeast Brazil, as well as near the São Francisco river mouth, due to the strong onshore winds there (Barbosa and Dominguez, 2004; Jimenez et al., 1999). In Argentina, dunes along the northern Buenos Aires coast have been stabilised from 1956 to 2013 due to the humid climate and human activities (urbanisation and afforestation) (Marcomini et al., 2017). Similar stabilisation has also occurred along the south coast of Buenos Aires since the 1940s (Amiotti et al., 2013). However, dunes on the Valdes Peninsula drylands are still active, with little change in the bare sand surface during 1969–2002, due to the intensive grazing there (Del Valle et al., 2008). The dunes of Ritoque in the central coast of Chile also experienced stabilisation during the last century, caused by planting, urbanisation and road construction (Nehren et al., 2016). The construction of artificial dunes by fencing, followed by marram grass planting for dune stabilisation was also conducted in some sites along the central Chilean coast (Castro, 1988).
5 Oceania
Extensive dunes occur along the coastline of Australia, covering an area of 2000 km2 (Short, 1988) (Figure S5, supplemental material). Dune stability has shown an increasing trend in Australia. This is well observed on Stradbroke Island, due to planting (Barr and McKenzie, 1976); Moreton Island and Fraser Island in Queensland, as a result of reduced windiness (Levin, 2011; Levin et al., 2017); the New South Wales coast due to dune stabilisation programs (NSW Department of Land and Water Conservation, 2001); and southern Australia, due to the decline of the rabbit population (Moulton et al., 2019). In New Zealand, the total area of mobile dunes was about 1300 km2 in the 1900s, with little change before the 1950s (Cockayne, 1911; Hilton, 2006). However, mobile dunes declined by 70% mainly during the second half of the 20th century as a result of marram grass planting and afforestation (Hilton, 2006). Bare sand area declined from 1294 km2 in the 1950s to 389 km2 in the 1990s, with the most significant reduction observed in regions of Manawatu (by 81.3%), Tasman (by 78.2%) and Northland (by 76.3%) (Hesp, 2001; Hilton, 2006). Another case can be found in Farewell Spit, where dunes showed a 16% reduction in the bare sand area during 1950–2004 due to the expansion of exotic vegetation (Tribe and Kennedy, 2010).
6 Asia
There are 133 dunes along the South Korean coast (MOE, 2001) (Figure S6, supplemental material). The Shinduri dune in the west coast is the largest dune in South Korea, which has largely been replaced by tourism urbanisation during the last few decades (Kim and Ewing, 2006; You et al., 2018). Kangneung dune on the east coast also showed a decrease in bare sand area by 61% in the 1910s–2000s, as a result of agriculture, land reclamation and the construction of power plants (Choi et al., 2007). Coastal dunes in China have been stabilised by planting and afforestation since the 1950s. For example, dune in Qinhuangdao, which is the largest coastal dune in China, has been gradually stabilised by afforestation, with an increase in forest cover from 8% in 1956 to 70% in the 2010s (Cai and Cai, 1983; Qi et al., 2004). More stabilised dunes can be found along the coast in Fujian (Chen and Zhu, 1996; Chen et al., 1992; Hu et al., 2013), Taiwan and Hainan Island (Figure S6) (Ho et al., 2017; Qu et al., 2013; Shih et al., 1994; Wu and Wu, 1987). Increases in coastal dune stability were observed in central Vietnam after a ban on the overuse of dune areas in 1995, followed by the tree plantation in 1997 (Hoang et al., 2010; Nehren et al., 2017). Stabilisation was also observed in southeast Vietnam as a result of land reclamation for agricultural use (Quang-Minh et al., 2010; Sharma, 1990), grass planting (Nehren et al., 2017; Tan Van et al., 2003) and tourism development (Geissler et al., 2011). Increase of dune stability in Parangtritis, Indonesia, was caused by tourism urbanisation since the 1980s (Nehren et al., 2016). In India, mobile coastal dunes in the Cauvery delta were largely stabilised in 1980 and converted into plantations and cultivation lands (Alappat et al., 2011). Cultivation, recreation and sand mining activities, as well as the coastal erosion, had caused a loss of active dunes in Ovari, southeast India, during 1999–2006 (Mujabar and Chandrasekar, 2012), as well as dunes in Vengurla in western India (Noujas et al., 2019). In contrast, mobilisation was observed on two dunes in southern India during 1999–2006, due to the high local wind speed (Mujabar and Chandrasekar, 2012).
V Drivers for dune mobility changes
1 Land-use change
Land-use change, including the abandonment/change of traditional farming practices (e.g. grazing and wood gathering), urbanisation and the development of tourist infrastructure, were quite common around the world during the last century. The loss of bare sand area caused by land use change is one of the main drivers for the increase in dune stability since the beginning of the 19th century.
1.1 Change in traditional farming
Traditional farming practices, such as grazing by domestic stock, marram cutting for mat/basket making, tree falling for firewood and cultivation, have existed in coastal dune areas for centuries. These activities often destroy the vegetation cover and lead to drifting sand and migrating dunes (Blanco et al., 2008; Jenks and Brake, 2001; Del Valle et al., 2008). For example, deforestation and land overuse led to extensive mobile dunes in Europe through the 1300s–1800s (Provoost et al., 2011). During the last century, many of these activities have been abandoned or restricted to minimise sand drift in many regions. For example, the removal of livestock around 1910 is considered to have initiated dune stabilisation in Oostvoorne (the Netherlands) (Van der Maarel et al., 1985). The decreased intensity of livestock grazing (cattle, horses and sheep) has led to vegetation expansion in the Netherlands (Van Dijk, 1992). Similar patterns of dune stabilisation caused by changes in grazing pressure have been identified in Iceland (Runólfsson, 1987), Israel (Kutiel, 2001), New Zealand (Tribe and Kennedy, 2010) and India (Alappat et al., 2011). In addition, reduced grazing pressure, caused by the eradication of rabbit populations, was cited to have contributed to an acceleration of vegetation succession and dune stability in northwestern Europe and southern Australia (Moulton et al., 2019; Provoost et al., 2011).
The ‘Mawasi’ agriculture system, where inter-dune areas were cultivated by digging the sand down to the water table for irrigation, has been a traditional practice in Israel for centuries (Levin and Ben-Dor, 2004). The ending of this agricultural practice is considered as one of the main contributors to the increase of dune stability along the Israeli coast (Tsoar and Zohar, 1985). The decline in agricultural land-use was one of the drivers for dune stabilisation in central Italy during the second half of the 20th century (Malavasi et al., 2013). By contrast, in some areas, coastal dunes were reclaimed (usually combined with the use of waste and fertiliser) for cultivation, due to the increasing populations and need for agricultural land (Gadgil and Ede, 1998; Hoang et al., 2010). This process was widespread in South Korea from 1910 to 1960 (Choi et al., 2007), as well as in the past three decades in Vietnam and China (Ho et al., 2017; Hoang et al., 2010; Neve et al., 2009; Shih et al., 1994).
The establishment of national parks also contributed to dune stabilisation by excluding or limiting human disturbance in dune systems. Doñana National Park in Spain, Prince Edward Island National Park in Canada and reserves and national parks in Israel are good examples of this (Levin and Ben-Dor, 2004; Mathew et al., 2010; Vallejo Villalta and Márquez Pérez, 2006).
1.2 Urbanisation and tourism
The urbanisation of coastal regions has been widespread around the world during the last century. Urban housing, the construction of industrial infrastructure (i.e. power plants), road networks and coastal protection constructions were also developed, which has led to a loss of dune habitats along the coast. For example, city construction and its accessory facilities have occupied more than 45% of the Israeli coast due to the population increase in the last century (Golik, 1997; Kutiel, 2001). In China, lagoons and dunes along the coast have disappeared due to residential, industrial and road constructions since the 1950s. Similar urban expansion and dune loss also took place in Belgium (Provoost and Van Landuyt, 2001; Van der Biest et al., 2017), Italy (Malavasi et al., 2013), Turkey (Kemal Sönmez et al., 2009), Mexico (Martínez et al., 2014), Brazil (Marcomini et al., 2017; Mendes and Giannini, 2015) and Chile (Nehren et al., 2016).
Coastal tourism, recreational pressures and the construction of tourist facilities, which intensified towards the end of the 19th century, have caused significant destruction of coastal dunes (Alcántara-Carrió and Alonso, 2002; Sytnik and Stecchi, 2015). A loss of coastal dune by 40–80% in Greece, by >95% in Morocco and by 85% in Indonesia was caused by tourism development (Doody, 2012; El Mrini et al., 2012; Nehren et al., 2016). Besides, resorts can also alter dune mobility by impacting the aeolian dynamics – for example, a shadow zone formed at the lee side of a resort in the Canary Islands, decreasing aeolian sediment transport and resulting in the increase of dune stability (García-Romero et al., 2019; Hernández-Cordero et al., 2018). Impact of tourism development can also be found in Catalan (Spain) (González-Villanueva et al., 2013), Qinhuangdao (China) (Qi et al., 2004) and Shinduri (South Korea) (Kim et al., 2014; You et al., 2018).
2 Dune stabilisation projects
Mobile dunes and drifting sand can invade forests and farms, and can bury roads and buildings (Sherman and Nordstrom, 1994). Over the past century, therefore, legislation (i.e. controls on overgrazing, woodcutting, trampling) has been established in many countries to minimise sand drift. For example, since the establishment of Israel in 1948, dunes were protected by law and were forbidden to disturb; while no laws/rules were enacted in Egypt, and human activities (such as grazing, wood gathering and human trampling) were still allowed during the same period (Kutiel et al., 2004; Levin and Ben-Dor, 2004). This led to a contrast of vegetated dunes (with dark colour) in the Israeli side and mobile dunes (with light colour) on the Egyptian side at the Israeli–Egyptian border (Figure 4). Similar legislation on dune protection was also established in many other countries, such as in Iceland (Runólfsson, 1987), the US (Godfrey and Godfrey, 1974) and New Zealand (Wendelken, 1974).

Vegetated dunes in Israeli side and mobile dunes in Egyptian side at the border area. The contrast illustrates the effect of legislation on dune management. (The image is referred from Yizhaq et al. (2007) and the arrow shows the prevailing wind direction.)
Along with the legislation, dune stabilisation projects were conducted worldwide to halt drift sand and mobile dunes, which played an important role in the dune stabilisation process. The main measures include grass planting and afforestation. Early attempts at dune stabilisation took place in western Europe, mainly by planting marram grass and maritime pines (Doody, 2012). Later, similar experience was well summarised in South Africa (Avis, 1989; Schwendiman, 1977), the US (Birkemeier et al., 1984; Reckendorf et al., 1985), Australia and New Zealand (Gadgil and Ede, 1998; Wendelken, 1974). In brief, fertilised sand-binding grasses (i.e. Ammophila arenaria) were used for the initial stabilisation stage, followed by the seeding of secondary permanent grasses and legumes when the sand movement slows down. Finally, woody plants and afforestation were introduced after the preliminary stabilisation was completed.
Dune stabilisation projects have been effective in stabilising mobile dunes, and similar efforts were conducted in many countries, including China (i.e. Casuarina equisetifolia) (Xie and Li, 2001; Xie et al., 2008), South Korea (i.e. Pinus thunbergii and Pinus densiflora) (Choi et al., 2013; You et al., 2018), Brazil (Pinus and Eucalyptus species) (Da Silva and Hesp, 2013), Chile (i.e. Ammophila arenaria, Pinus radiata, Pinus halepensis, Casuarina cunninghamiana, as well as Eucaplytus and Acacia species) (Nehren et al., 2016) and Vietnam (such as vetiver grass) (Tan Van et al., 2003).
One problem with these dune stabilisation projects is that exotic species were widely used, many of which are invasive. This is an important driver for dune stabilisation in some regions, but, on the other hand, it also has adverse impacts on the indigenous species, mainly causing a loss of native species. For instance, two Ammophila species (A. arenaria and A. breviliigulata) had colonised the entire Pacific coast within a few decades, resulting in dune stabilisation (Hacker et al., 2012; Heathfield and Walker, 2011). However, the invasion resulted in the decline of several endemic dune plants along the northwest America coast, such as pink sand verbena (Abronia umbellata ssp. Breviflora) (Hacker et al., 2012). Similar invasions of marram grass (A. arenaria), sea-wheat grass (Thinopyrum junceiforme) and bitou bush (Chrysanthemoides monilifera ssp. rotundata) were reported in New Zealand and Australia (Hilton et al., 2006; Thomas et al., 2006). Due to the invasion of Acacia species, the soil nutrient concentrations increase by two-fold on the Robberg Peninsula, South Africa, leading to a reduction of the indigenous dune species, which had adapted to a nutrient-impoverished environment (Musil and Midgley, 1990). Exotic invasive black pines (P. thunbergii), which have been introduced to South Korea for coastal dune stabilisation since the 20th century, and are now the most dominant species on the coast there, covering approximately 95% of the total dunes (Kim, 2005; MOE, 2001).
Combined with planting projects, sand-trapping fences had also been used to assist dune stabilisation because they can cut off the supply of sand to inland areas, as well as break the force of winds at ground level (Pelletier et al., 2009; Wendelken, 1974). For example, bamboo fences worked very well to trap the sand in Changle dunes along the coast of China (Huang and Yim, 2014; Lin and Liou, 2013; Liu, 2007). Similar fencing was also widely used in many cases in Spain for dune stabilisation (Gómez-Pina et al., 2002). More than 1000 km of sand fencing was installed on the Jorkey’s Ridge dunefield in the US to slow dune migration (Mitasova et al., 2005). More cases can be found in New Zealand (Wendelken, 1974), Egypt (El Fishawi and El Askary, 1981), South Africa (Avis, 1995) and Japan (Cong, 1991; Zhu and Yang, 1987) and India (Mascarenhas, 2002, 2008).
3 Sediment decline
Sediment supply from the coast is the sand source for the formation and growth of coastal dunes (Davidson-Arnott and Law, 1996). Change in sediment supply can modify dune mobility. Specifically, high sand supply can bury vegetation, resulting in drifting sand and mobile dunes, while low sand input can encourage vegetation survival, leading to an increase in dune stability (Pye and Blott, 2017).
Fluvial sediments are also an important sand source for coastal dunes – given that many dunes are usually associated with the estuaries/ocean outlets of large rivers or streams – such as the dunes in Oregon, US, South Africa (Avis, 1995), Copiapó region, Chile (Paskoff et al., 2010), southern Mozambique coast (Miguel and Castro, 2018) and most coastal dunes in China (Cai and Cai, 1983; Dong, 2006; Schwendiman, 1977). It is reported that there is a global decline of sediment supply from rivers to the ocean, especially since the 1950s, caused by the sediment retention of reservoirs (Syvitski et al., 2005; Syvitski and Milliman, 2007; Vörösmarty et al., 2003). It is estimated that there are more than 48,000 large dams globally, with more than 2000 large dams under construction (Syvitski and Kettner, 2011). Large reservoirs, on average, can trap more than 90% of the river sediment to the coast. For example, the construction of the Low and High Aswan Dam (in 1902 and 1964, respectively) in Egypt entrapped almost all (>98%) sediment behind the dam (Stanley and Warne, 1993), which contributed to dune stabilisation along the eastern Egyptian coast and the Israeli coast (Kutiel, 2001; Muhs et al., 2013). A similar impact of hydraulic projects on dune stabilisation also occurred in Turkey (Berberoğlu et al., 2003), China (Qi et al., 2004) and Vietnam (Sharma, 1992), where large amounts of riverine sediment were trapped behind the dams.
Coastal structures such as the dykes, groynes and breakwaters can also modify the littoral drift (Nordstrom, 1994). Alongshore sediment deposition usually occurs in the updrift side of the structures, while erosion takes place in the downdrift side, leading to sand deficient and dune stabilisation on the back of the beach (Martínez et al., 2019). Dune stabilisation, partly facilitated by the coastal structures, was reported in the Netherlands (Aagaard et al., 2007), Spain (Rodríguez-Ramírez et al., 2008), Brazil (Martinho et al., 2010) and Mexico (Martínez et al., 2019) with a declined alongshore sediment transport.
4 Changes in climate
Climate was cited as an important driver for dune stabilisation in different parts of the world. In most cases, it is hard to quantify climate impact on dune evolution since climatic factors work in combination with human activities to drive dune stabilisation. For example, most of the Danish dunes were active, and plantation and afforestation were not successful around 1820 when the climate was very stormy (Clemmensen et al., 2014). However, successful afforestation and dune stabilisation were achieved when the storminess decreased around 1900, indicating the impact of wind on dune mobility (Clemmensen and Murray, 2006; Clemmensen et al., 2014). Similar results – that is, that a combination of change in climate (i.e. less windy, more humid and warmer) and human activities (i.e. removal of grazing and afforestation) led to increases in vegetation and dune stability – were reported for dunes in the UK (Pye et al., 2014; Pye and Blott, 2017; Rhind et al., 2001), France (Ruz et al., 2005), Spain (González-Villanueva et al., 2013), Israel (Levin and Ben-Dor, 2004), India (Jayangondaperumal et al., 2012; Kunz et al., 2010) and China (Hu et al., 2013) during the past century.
Climate is also suggested to be the dominant driver for increases in dune stability in areas where human activities were limited or rare. For example, most dunes in Ireland are isolated and free of human activities. Dune stabilisation there has been coincident with increased temperature and longer growing season since the 1960s, indicating that climate rather than human intervention is the controlling factor for dune mobility change (Jackson and Cooper, 2011). Similarly, dune stabilisation on some islands in Brazil was linked to changes in rainfall and wind speed (Da Silva and Hesp, 2013). Shrub encroachment on Hog Island was caused by local climate warming (Huang et al., 2018). More dune sites with rare human intervention around the globe show a global ‘greening’, driven by a combination of climate change (increasing rainfall, warming and reduced windiness) and increased atmospheric nutrients deposition (Jackson et al., 2019).
Other climatic factors, such as storms and sea/water level change, are also important drivers for dune mobility change. The frequency and magnitude of storms controls the spatial-temporal pattern of foredune erosion and the sediment transport process (Davidson-Arnott et al., 2018). Dune erosion caused by storms occurs within hours or days, but the recovery can take years to decades (Houser et al., 2015). The most recent sand invasion during the Little Ice Age in Europe was caused by an increase in North Atlantic storminess (Clarke and Rendell, 2009). Dune erosion and accretion in the UK was associated with the frequency and magnitude of storms (Pye and Blott, 2008). The overall decline of storminess in the 20th century contributed to dune stabilisation in Northern Europe (Provoost et al., 2011). Dune migration at southern Lake Michigan was linked to the frequency and intensity of the storm, rising lake level, higher temperature and moisture (Kilibarda and Shillinglaw, 2015). An intense storm in 1932 caused foredune destruction and a 300–600-m inland migration of transgressive dunes in east Canada, and the recovery took about four decades (Mathew et al., 2010). Similarly, storms caused coastal dune erosion in southern USA as well as in northern France, and post-storm recovery took about 10 years (Héquette et al., 2019; Houser et al., 2015). Dune stabilisation on Fraser Island in Australia was coincident with increased cyclone frequency and intensity (Levin, 2011).
Future sea-level rise will inevitably cause marine flooding, coast erosion and constrain the area for dune plants (Feagin et al., 2005; Miller et al., 2010; Nicholls and Cazenave, 2010). Mobile dunes could be vulnerable to coastal erosion and flooding. However, vegetation on fixed dunes can reduce scarp retreat, wave run up erosion and dune structure failure (Delgado-Fernandez et al., 2019b; Feagin et al., 2019; Maximiliano-Córdova et al., 2019; Sigren et al., 2014). Furthermore, it was also suggested that the effects of sea-level rise may be obscure in developed areas since human efforts may protect and maintain a stable resource base in some degree (Nordstrom, 1994).
Similar to the global ‘greening’ observed by Jackson et al. (2019), a global increase in dune stability was found in this study, although we considered a longer time span and a larger number of dune sites with either considerable or rare/limited human interventions. Jackson et al. (2019) elucidated the impact of change in climate and atmospheric deposition on dune mobility, by considering dunes only from the natural coast with rare human interventions. However, human activities are considered to be the main factor driving shifts in dune mobility for most dunes in this study, although climate has also helped to promote dune stabilisation (Figure 5).

Main drivers for dune stabilisation in this study. The increase in stability of most dunes is dominated by human activities, including land-use change, stabilisation project and sediment decline caused by hydraulic constructions. Change in climate also plays an important role in dune stabilisation, especially for dune sites at coasts with limited human impact. However, human intervention can also reverse dune stability through vegetation removal to rejuvenate dune mobility, which has been conducted as trials or on small scales in a few countries during the past three decades.
An unintended consequence of dune stabilisation is that a loss of biodiversity has been found in the over-stabilised dune areas during the past three decades. This has led to deliberate de-stabilisation programs (e.g. mechanical removal, grazing, herbicide-based de-vegetation, hand pulling) to ‘rejuvenate’ the variability and biodiversity of dune habitats (Figure 5) (Darke et al., 2016; Konlechner et al., 2014; Pye et al., 2014). Such de-stabilisation measures have been conducted in different dune systems, including Kennemerland dune in the Netherlands (Arens et al., 2004, 2013), Nizzanim dunes in Israel (Bar, 2013), Steward Island in New Zealand (Hilton and Konlechner, 2010; Konlechner et al., 2014, 2015), Lanphere and Ma-le’l dunes in the northern California (Pickart, 2013) and Wickaninnish dunes in Canada (Darke et al., 2013, 2016). Increased dune variety and dune reactivation were observed in these dune sites after the projects. However, these ‘rejuvenation’ efforts are still quite controversial, since stabilisation could be the natural evolutionary processes taking place in the dunefields (Delgado-Fernandez et al., 2019b); besides, these deliberate de-vegetation projects were started at a small scale. There could be a re-activation trend in these sites in the future; however, a longer time is required to examine whether these projects can naturally re-mobilise the dunes on a larger scale or in the long term. In other cases where specific boundary conditions are dominant, such as the unlimited sediment supply in the southern Spain and strong winds in south India and northeast Brazil, dunes are continuing their mobility trends (Mujabar and Chandrasekar, 2012; Navarro-Pons et al., 2016; Tsoar et al., 2009).
VI Conclusions
This review shows a similar global ‘greening’ or stabilisation trend of coastal dunes, as reported by Jackson et al. (2019), by utilising a larger temporal window and more dune sites. However, most (82% or 144 out of 176) of these dune sites are impacted by human intervention, which, on the one hand, may reflect the significance of human alteration on the coast (especially in Euro-Mediterranean areas) and, on the other hand, may also cause a bias of human impact due to oversampling in these areas.
The increase in dune stability was mainly in combination with an increase in either vegetation cover or urbanisation area during the past century. The main drivers were: (a) land-use change, such as change of traditional farming, urbanisation expansion and tourism development; (b) dune stabilisation projects; (c) sediment decline caused by riverine and coastal constructs; and (d) change in climate and storminess. From all these factors, human activities are considered to be the dominant driver for dune stabilisation, while climate and storms also contribute to the increase in dune stability, especially for those remote sites with rare human intervention.
In the context of coastal urbanisation, as well as global warming and associated sea-level rise, coastlines are more vulnerable to overdevelopment and dune erosion. Mobile dunes are suggested to have a higher ecological diversity and species richness, while stabilised dunes could be more tolerant to future sea-level rise. Human intervention plays a vital role in altering dune mobility (both stabilisation and re-mobilisation). Therefore, more scientific attention and cautious measures would be needed in the near future to maintain or reverse dune mobility, as well as to maintain a balance between the services that coastal dunes can provide and the hazards they can lead to.
Supplemental material
Figure_Supplemental_Material - Coastal dune mobility over the past century: A global review
Figure_Supplemental_Material for Coastal dune mobility over the past century: A global review by Jinjuan Gao, David M. Kennedy and Teresa M. Konlechner in Progress in Physical Geography: Earth and Environment
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
