Introduction

People inhabiting the arid and semi-arid zones of Earth remain among the most vulnerable populations to environmental changes. Rural communities are particularly vulnerable, as they heavily rely on natural resources for their livelihoods (Borderon et al. 2019, p. 521). This dependency is particularly fragile on the borders of continuous deserts, with limited water resources and intensifying land degradation (including desertification). Limited agricultural development possibilities have led people to seek various solutions in the unceasing struggle against environmental limitations. Migration,Footnote 1 often a last resort, is increasingly seen as an adaptation strategy (Schraven et al. 2021).

More studies use a comparative approach to examine the relations between the environment and migration in arid and semi-arid areas (Obokata et al. 2014; Neumann and Hermans 2017; CLIMIG 2018). However, discussions about the environmental causality of migration are often conducted in isolation from the social, political, and economic factors. Despite that, there can still be a profound understanding of the complex relationships between the factors that strengthen or mitigate environmental pressure on human activities and trigger migration (Entwisle 2021; Hermans and McLeman 2021, IPCC 2022).

This paper explores the interplay of environmental and non-environmental factors influencing mobility in three North African oases: Mhamid in Morocco and El Faouar and El Ferch in Tunisia. It reveals how water scarcity, agricultural challenges, and socio-economic conditions interact to shape mobility patterns. By examining these factors though a comparative lens, this paper contributes to a nuanced understanding of the mobility-environment nexus and highlights diverse local adaptation strategies to changing conditions.

The analysis is based on mixed-method research, including field visits, questionnaires, in-depth interviews with heads of households and local actors, and basic environmental research conducted in the aforementioned oases.

The role of the environment in human mobility in arid areas—a review and conceptual gap

The investigated regions have traditionally relied on agriculture; therefore, a closer look at the relations between the environment, agriculture, and mobility is necessary. Up to now, many scholars have reflected on the arguments of De Haas (2001, p. 37), who claimed that “sufficient income is a necessary, but not exclusive, condition for investment in agriculture”. He indicated that environmental factors, such as climate and its variability and land and water resources, can be incentives or discourage agricultural investment. Contrary to Afifi and Warner (2008, p. 16), who wrote that the most significant environmental factors (for migration) are soil quality and the availability of suitable water, De Haas (2001, p. 37) showed, with the example of Moroccan and Tunisian oases, that scarcity of land resources can be a much more significant barrier. Nonetheless, he admitted that insufficient water resources may lead to agricultural decline, limitation of income, and finally, when resources for new pumping or state support are lacking, abandonment of households. He also indicated that the quality of soil and water resources is not as important for agricultural development as long as it is acceptable for traditionally developed coping solutions. Further, Henry et al. (2004, p. 417) confirmed a more significant role for slow-acting processes, such as land degradation, than for episodic events, giving the example of drought. They claimed that “land degradation progressively makes livelihood strategies based on farming unsustainable”. Also, the Groundswell II Report highlighted that the increased movement of people to cities is due to climate-related challenges such as water scarcity, declining crop productivity, and sea-level rise (Clement et al. 2021). In such a situation, the role of insufficient and/or low-quality water resources and land degradation that influences crop productivity becomes relevant, as increasing research shows that significant deterioration of water quality limits agricultural development—clearly visible in arid and semi-arid areas (Kadri and Ranst, 2002; Habib 2013; Dłużewski et al. 2017; Stoler et al. 2021, IPCC 2022).

Natural resource-related drivers, however, are just some of the many factors infused in mobility decisions (Kniveton et al. 2008; Foresight Migration and Global Environmental Change, 2011). Research on the environment-mobility nexus widely acknowledges that migration is multi-dimensional and multi-causal. Environmental changes often affect mobility indirectly by influencing other drivers of mobility, such as socio-demographic, economic, and political factors (e.g. Afifi et al. 2012); social networks (e.g. Simatele and Simatele 2015); household characteristics (e.g. Kubik and Maurel 2016); or gender vulnerability and heritage systems (e.g. Ou-Salah et al. 2023). Particularly relevant in the context of the areas under discussion is the relationship between environmental and economic factors. Neuman and Hermans (2017) found that economic-social motivations consist of 80% of all identified drivers from investigated case studies and that only 11% of the drivers were linked to demographic and/or environmental conditions as a direct cause, and most of those studies did not explore the causation between them. Such approaches limit the opportunities to understand the complex relationships between the various factors. An analysis of empirical evidence carried out by Borderon et al. (2019, p. 518) indicated that environmental degradation via a reduction of economic means often pushes people to resort to resources outside the local territory (migration as an adaptation strategy that allows for diversification of sources of income) (i.e. Konseiga 2007; Neumann et al. 2015, McLeman et al. 2021, Sakdapolrak et al. 2024). Afifi (2011, p. 116), on the other hand, showed that the primary causes of economic causation in Niger were declining crop yields and cattle deaths due to water shortages. Further, research by Mounkaïla (2002), also carried out in Niger, indicated two migration practices, each revealing a specific household and/or family strategy to either fight for the heritage (stay in the region) or abandon it (permanent migration). In the first practice, according to the New Economics of Labour Migration (NELM), migration is part of a diversification of sources of income and survival strategy in the region of origin, which over time increases dependence on migration resources (Stark and Bloom 1985). The second tends towards the permanent migration of some people, and in this practice, migration serves as a bridge to a new anchorage in a town or other rural area. Both responses were confirmed by studies of other regions (e.g. Penning-Rowsell et al. 2013; Williams and Gray 2020) and are covered by the International Organization for Migration’s definition of environmental migration (Perruchoud Redpath-Cross 2011).

The close interplay between the environmental and non-environmental factors that drive migration makes the study of environmentally induced migration extremely difficult. Among the attempts to conceptualise the relationship between both of them in out-mobility decisions, the predominant focus is on typologies of migrants: “voluntary”, “forced”, etc. Scholars have created multiple typologies of migration motivations (Suhrke 1993; e.g. Bates 2002; Castles 2002; Hunter 2005), and Stojanov et al. (2014) prepared a comprehensive overview of them. Particularly noteworthy in this context is the typology proposed by Renaud et al. (2011) that allows for a qualitative differentiation of the types of environmental migrants and demonstrates how the relationship between different factors in various contexts may create different types of migration. These authors identified “emergency environmental migrants”, “environmentally forced migrants”, “environmentally motivated migrants”, and migrants who are not described as environmental but who move away from an area affected by environmental degradation although that does not dominate their migration decisions. Such a focus on migrant categories, however, does not say much about the mobility process itself and is often a reflection of what Sakdapolrak et al. (2016, p. 86) described as “a lack of recognition of the full complexity of migration”. Promising in this context is the typology of interactions created by Morrissey (2013), which he classifies as “additive”, “enabling”, “vulnerability”, and “barrier effects” that ultimately impact the mobility-environment connection. Still, as Freeman (2017, p. 355) points out, this focus on interactions ignores the multitude of non-environmental factors that shape mobility in ways that are not obviously linked to factors rooted in the biophysical environment.

Few researchers have pursued conceptual solutions beyond migrant typologies. Black et al. (2011) proposed a conceptual framework for understanding and assessing the impact of environmental change on other mobility factors, such as political, demographic, economic, and social, while noting the complexity of the relationships between them. While this is a significant attempt to better conceptualise the relationship between environmental and non-environmental factors (Morrissey 2013, p. 1502), it still treats the environment as a driver of mobility but provides no means to conceptualise it. Further, Neumann et al. (2015) conceptualised environmental drivers of migration in drylands and applied a cluster analysis for analysing and mapping these drivers at the global scale. Finally, Zickgraf et al. (2022) made a further attempt to frame the relationship between mobility drivers, particularly in the context of pressures on natural resources resulting from environmental change. They point out that the multitude of linkages and feedback between natural resources and migration lend themselves to a complex adaptive (sub)system framing within larger socio-ecological systems. This systems approach allows us to better understand the dynamics of the relationship between the factors that generate mobility. At the same time, it fits into a research agenda that goes beyond linear theories of environmental “push” or economic “pull” towards one with a greater appreciation of the context and non-linear relationships at all levels (Hunter et al. 2015; McLeman and Gemenne 2018).

The prevailing view of the conceptualisation of the environment-mobility nexus is that migration decisions are influenced by many variables and that mobility is context-dependent and highly complex. However, this complexity “seems to be a major conceptual and methodological obstacle” to research in this domain, where the conclusion that migration is complex represents a profoundly unsatisfying result (Land, Romankiewicz and van der Geest, 2018, p. 170), reflecting a significant gap in research on the subject. Thus, there is a clear need for a better conceptual and methodological framework to deal with this complexity. This text is an attempt to address these deficiencies by analysing the multidimensional factors that cause different types of mobility in similar, yet different, contexts.

Methodology

This paper relies on research carried out between 2015 and 2019. The analysis began with extensive desk research on social and environmental contexts, followed by a mixed-method approach combining qualitative and quantitative data. The team conducted field research, performing non-participatory observation, and compiled photo documentation to understand the relationship between environmental constraints and human responses.Footnote 2

Quantitative research

The questionnaire was aimed at providing descriptive and contextual data and indicating the best cases for in-depth interviews. The sample for the survey was chosen through random walks. In Mhamid, Morocco, 120 micro-questionnaires were collected, followed by an extended questionnaire interview with 48 heads of households. In Tunisia, the micro-questionnaire phase was merged with the extended one, and in El Faouar and El Ferch, 120 and 87 households were investigated, respectively. The questionnaire was in French; optionally translated into darija (spoken Arabic, which differs from one country to another). The descriptive quantitative analysis was conducted with the use of SPSS and Excel.

Qualitative research

The qualitative study was elaborated to understand the process, focusing on the relationships between the environment, agriculture, mobility, and adaptation strategies. In-depth interviews were carried out with the heads of households until the level of saturation was reached, which was after 12 households in Mhamid, 17 in El Faouar, and 16 in El Ferch. The interview questions concerned the perception of the quality of the natural environment, the impact of individual environmental factors on agriculture and income, alternative sources of income of respondents, mobility of family members, and whether and how this was related to limitations and degradation of the natural environment. The interviews were conducted in Arabic (darija) and/or in French with an interpreter. After consent, the interviews were recorded, translated, and transcribed into English and subsequently analysed using AtlasTi. When permission to record was not granted, precise notes were taken and analysed similarly.

Integration of quantitative and qualitative data

The integration of quantitative and qualitative data was essential to provide a comprehensive analysis of the similarities and differences between the case studies. Quantitative data provided a broad overview and contextual background, allowing us to identify patterns and correlations. The qualitative data added depth and detail, illustrating how individual experiences and perceptions aligned with or diverged from these patterns.

By comparing the quantitative survey results with the qualitative interview findings, we could identify common themes and unique factors across the three oases. For instance, while all of them faced challenges related to water scarcity, the specific impacts and adaptive strategies varied. In Mhamid, the focus was on historical changes due to a dam’s construction, while in El Faouar, the emphasis was on the increasing demand for private wells. In El Ferch, the historical pattern of international migration played a significant role.

This mixed-method approach allowed the research team to highlight the similarities and differences across the case studies, providing a nuanced understanding of how various factors interact to influence mobility decisions.

Characteristics of the research areas: water scarcity and the socio-ecological transition

The research was carried out on three oases located at the edge of the Sahara Desert: Mhamid in Southern Morocco, El Faouar in Middle Tunisia, and El Ferch in Southern Tunisia (Fig. 1). In all of the investigated areas, we can observe three important elements of the subject of study: degradation of the environment, limitations imposed by the environment on human activities and agriculture, and limitations imposed by the same on human mobility (both internal and international).

Fig. 1
figure 1

Location of the field studies in Morocco (left) and Tunisia (right) (prepared by P. Jaczewski)

Full size image

Looking at the three areas, we find that they are very similar from both a geographical and social point of view. All of them are situated on a border with continuous desert, on flat terrain with small oases, far from the centre of the country. All three are located in hot desert climates with average yearly precipitation between 90 and 110 mm and rainy seasons between September and October and in March. The maximum temperature, usually in July or August, reaches over 40 °C, and the minimum temperature falls below 6 °C in January, giving an average yearly temperature of slightly over 20 °C (Climate-Data.org 2021). Also, soil-quality conditions are comparable. In all three oases, we can find sandy and fertile soils, where low water quality causes salinisation, which significantly influences the quality of crops. Agriculture and breeding are the main income activities.

Like all countries in North Africa, these oases faced a demographic boom in the latter twentieth century, which resulted in significant increases in population (Table 1), putting increased pressure on meeting the growing needs of the inhabitants, with some decrease later on. In Mhamid, the population decreased significantly, from more than 9000 in the 1970s (Ait Hamza et al. 2009) to 5857 in 2014 (RGPH 2019) due to migration. The El Faouar délégation is the largest in the Kebili governorate,Footnote 3 with eight villages over an area of 9,671.3 km2, and its population increased 20-fold between 1970 (about 1000) and 2014 (over 19,000) (INS, 2016a; Sobczak-Szelc and Fekih 2020). El Ferch Oasis is one of the sectors in the Ghomrassen délégation in the Tataouine governorate. The sector covers an area of 2990 ha, with 295 households and 1307 inhabitants (INS 2016b). Although the population of Tataouine more than doubled over four decades, from 70,212 inhabitants in 1975 to 150,000 in 2018 (INS estimates), the number of inhabitants of both the Ghomrassen délégation and El Ferch sector has decreased, especially since 1994 (Table 1).

Table 1 Number of inhabitants and households in Mhamid, El Faouar, and El Ferch between 1970 and 2014
Full size table

What makes Mhamid and El Faouar different from El Ferch from an environmental point of view are aeolian processes and the advancement of aeolian forms that encroach on fields and houses and hamper local human activity (Sobczak 2008; Ait Hamza et al. 2009; Sobczak-Szelc and Fekih 2020). The most significant difference lies, however, in the availability of suitable water resources.Footnote 4 Until the 1970s, Mhamid depended on underground alluvial waters. However, the situation changed after the Mansour Eddahbi Dam was built near Ouarzazate city in 1972. Since then, as the last of the six oases located in Draa Valley, Mhamid received water supplies from the dam for irrigation four to five times a year through a system that includes the riverbed, smaller dams, and channels. In practice, it was significantly less than assumed when building the dam and much less than when the Draa River had a periodic character (Sobczak 2012; Karmaoui 2015). Without the delivery of freshwater supplies in the alluvia, the existing water became saline and the level decreased. Lately, wells have been deepened due to the dropping groundwater levels, with no significant improvement in regard to salinity. As far back as 1967, the depth of the groundwater table varied from 2 to 8 m (Pletsch 1971); in 2005, it was 3 to 12 m, and in 2015, it ranged from 6.5 to 17 m under the terrain surface (field observations). Further climate scenarios show that the available groundwater resources will decrease significantly up to 2029 and will not meet the growing demand for irrigation, especially during years of drought.

In El Faouar, where surface water resources are not available, agriculture relies on artesian and subartesian waters extracted from state wells and delivered to the fields through channels. The national wells dug in the early 1960s turned out to be insufficient for the growing needs of new fields, and through overexploitation, the quality and quantity of available water have been deteriorating (Louhichi et al. 2000; Latos et al. 2018). Only a limited number of inhabitants have access to water from wells drilled down to hot springs (Mekki et al. 2013), which after cooling is used for irrigation. Therefore, others are pushed to look for different sources of water for irrigation.

Finally, in El Ferch, we can observe both rain-fed and irrigated agriculture. The former relies on rainy seasons and is focused on olive trees, while the latter uses groundwater, which allows for the cultivation of very profitable vegetables. Groundwater resources also face changes in quality and quantity. Like in Mhamid, the first underground wells turned saline in the 1970s. Today, fields are irrigated either from household wells dug 15 m deep or from the water line from Guermessa mountain, connected to some areas of El Ferch in 2004, or they rely on seasonal rains.

In all three oases, insufficient quality and/or quantity of water resources led to a decrease in the quality and quantity of crops, and the lack of possibilities to increase arable area limited agricultural development and income from this sector for the residents’ descendants. Such a situation had economic consequences and pushed inhabitants of all three oases to develop activities in addition to agriculture. In Mhamid and El Faouar, these were mainly tourism and service in the army. In El Ferch, the main additional activities were related to trade, sweet shops, construction, and other services. Most relevant, however, was that the worsening economic situation increased mobility (Sobczak-Szelc 2023; Sobczak-Szelc and Fekih 2020), especially as changing local and global geopolitical and economic conditions resulted in challenges in maintaining additional sources of income in all three areas. In Mhamid and El Faouar, tourism has become stagnant, while both El Faouar and El Ferch are confronted with Tunisia’s weak economy in which the unemployment rate in the fourth quarter of 2020 reached 17.4% (INS 2021). The absence of industry to offer new jobs for the growing number of unemployed youth drives people to leave their areas of origin. This is common to support the families left behind (Karmaoui 2015) and indeed, all three oases have strong emigration traditions. The inhabitants of Mhamid and El Faouar have always been nomads and merchants, and in the region of the Draa Valley, 65% of households depend on remittances from migrants (Rössler et al. 2010). El Ferch is in a region with traditional emigration to Europe established already in the 1960s.

To explore how the constellation of relations between environmental, demographic, economic, social, political, cultural, and historical factors influence mobility in rural areas and how far the interactions play out in a different context, the narratives of the three cases are juxtaposed below. The factor triggering the process ending in a mobility decision by a significant number of inhabitants of the three case study areas is different for each: in Mhamid, it is the degradation of the environment due to development; in El Faouar, it is limited access to resources as an effect of population growth; and in El Ferch, it is a mix of the two. These differences are extended depending on the coexisting social factors, including local aspirations, economic reasons, and other phenomena.

The complex relation between environment and human mobility: an empirical perspective

Mhamid: mobility to flee a changed environment

Mhamid is a traditional oasis on a former trade route to Timbuktu dating back to 3000 BC. It changed significantly after the construction of the Mansour Eddahbi Dam, which influenced not only the flow of water in the nearby riverbed, but the entire socio-ecological system. The increased irregularities in the flow of water in the riverbed have lowered the level of the groundwater. Also, climate variability has influenced the quantity of water resources delivered to the reservoir behind Mansour Eddahbi Dam, which in dry years postpones the initial flow of water from the dam, providing insufficient time for vegetation growth in the oases downstream. The lowering of water supplies has lowered the level of alluvial waters, which has influenced the degradation of vegetation cover and launched aeolian processes. Irregularities of water flows in the riverbed have influenced not only the vegetation cover but also the quality of alluvial groundwater, making them more saline. Irrigation of fields with mineralised waters has increased the salinity of soil and led to its further degradation. All of this has caused significant limitations for agriculture, mostly visible in income, which for some oasis inhabitants was already not sufficient to meet the basic needs. The limitations on agriculture have resulted in the mobility of farmers or household members, who move to other regions of Morocco (Fig. 2). That was the case of the brother of one head of a household who left at that time. In his statement, he noted that:

After the dam construction, there were years with rain and years with drought. (…) There was a drought for five years. In 1977, I left. There was no river, no water, no rain. I had to leave to live. (Mhamid HB141)

Fig. 2
figure 2

A simplified model presenting the changes in the socio-ecological system of Mhamid Oasis due to the construction of the Mansour Eddahbi Dam

Full size image

This quote illustrates the direct impact of the changes in water availability on mobility decisions, a relationship further supported by survey data and environmental observations.

Since 1995, the role of agriculture has decreased systematically and has been limited to a sentimental role, only for household needs. In satellite photos, we can observe a decrease in cultivated areas, from 4.89 km2 in 1995 to 3.85 km2 in 2005, and to 1.94 km2 in 2015. These findings from satellite imagery correlate with the interview statements, confirming the casual link between environmental degradation and reduced agricultural viability.

In the second half of the twentieth century, similar to other regions of North Africa, Mhamid faced an increase in the number of inhabitants. This change was not only the result of a higher birth rate (2.9 in 2004) but also from settlement by nomads, who were encouraged by the state and education system to remain at the oasis. Demography as a factor influencing mobility seems, however, to be less crucial in this case, because without the negative changes in the environment, the inhabitants would be able to maintain and even develop their heritage.

In the 1970s, mobility was just one of the options that, over time, gained importance. At the beginning, the inhabitants of Mhamid were looking for coping strategies. Facing limited possibilities for adaptation to the consequences of environmental change on agriculture and insufficient income from that branch of the economy, people were forced to look for sources of income outside it. For a long time, the most significant additional sources of income were the army, tourism, and remittances from those who had left for the big cities of Morocco. Back in 2009, according to the inhabitants, each family had at least one member working in tourism or the army (Sobczak 2012). With changes in the global economy, first after the 2008 economic crisis and then the Arab Spring in 2011, income from tourism in particular dropped (see Fig. 2, grey bar). In the survey, conducted in 2016, over 90% of investigated households reported at least one source of income other than from agriculture. Moreover, over 54% indicated two or more non-agricultural sources of income, and these came from employment in the army (68%), remittances (65%), and other services (23%). Employment in the tourism and construction sectors was only at 17% each. The survey data, combined with qualitative insights, demonstrate how economic changes influenced by environmental degradation have led to diversification of income sources and increased mobility.

Currently, despite additional sources of income, people are not willing to stay in the region. They want to leave not only agriculture but also Mhamid and southern Morocco, as explained by the father of two sons:

You can’t work anymore in agriculture due to climate change. One year is good and then you have 5-10 years of drought. That’s why my sons won’t stay here. (…) In Mhamid, much water is rare. Abandonment of work due to one year with enough water is uneconomical. My sons decided that they won’t stay here because, if it is good here for two years it is good, but then there is drought and you have no work. (HH44)

This statement underscores the causal relationship between climate change, water scarcity, and migration decisions, highlighting the complex interplay of factors influencing mobility.

The desire for mobility is additionally strengthened by the lack of possibilities for youth to fulfil their aspirations locally for having new things, such as phones or clothes. The decrease in tourism significantly limited options for additional income as well. In Mhamid, we can observe a shift from mobility that is dominated by environmental drivers, particularly for those who relied on agriculture, towards mobility that is environmentally motivated, but where the needs and aspirations of the youth gain significance.

El Faouar: mobility to overcome the limitations of the environment

After 1960 when El Faouar was established, it experienced both population growth and environmental changes. Increased numbers of descendants caused land fragmentation and a decrease in average household income. Cultivation of new terrain soon revealed insufficient water supplies. State wells started providing water every 20 days instead of every 12 or 14, causing the quality and quantity of date crops to deteriorate. The need for new wells and resources for new drilling became obvious, but the inhabitants of El Faouar became caught in a vicious circle (Fig. 3): as the state did not supply new water sources, the people started to act on their own by digging private wells to irrigate new fields, but this lowered the groundwater and even caused the private wells to dry up. The inhabitants again faced the challenge of insufficient water for irrigation and a demand for new wells. Nonetheless, the agricultural area, unlike in Mhamid, has increased with each passing year. In 1985, the total area of the fields of the three oases belonging to El Faouar commune, namely El Faouar, Sabria, and Ghidma, was 3.82 km2. In 1995, it was 6.54 km2; in 2005, it was 14.09 km2; and in 2015, it reached 29.18 km2. This shows a strong determination to develop agriculture in El Faouar, which is however limited by insufficient water resources, as explained by the head of one household: It is obvious for us to develop agriculture but the problem is with water. (…) If there is water, we don’t need to have other work. (H73).

Fig. 3
figure 3

A simplified model presenting the relation between human activity and changes in the environment in El Faouar Oasis

Full size image

The households there have found different ways of adapting to the limited water resources. Some of them try to cooperate with neighbours, plan for the future, and share expenses. Some look for different sources of income, as the head of a household in Sabria emphasised: You have to mix the sources of income. You have to do this way to live here. (H121, male, Sabria). In El Ferch, almost 70% of investigated households indicated at least one source of income outside of agriculture. Slightly over 20% indicated two or more additional sources. This share is significantly lower than in Mhamid, but it shows that agriculture can hardly be the main source of income for El Faouar. The other main sources of income mostly come from construction, the army (El Faouar is near the Algerian border), casual work (usually working in the fields of other farmers), trade, and remittances. Mobility (both internal and external) plays an important role here. Tourism has lost its significance as well and hopes for the future were curbed by the COVID-19 pandemic.

Those acting on their own look for resources for new wells through state loans or mobility. The former are available only to those employed in state positions, while the latter is only for those who can send at least one member to the big cities of Tunisia or abroad.

Having a migrant member is a factor that significantly distinguishes households. Among the 80% of those with at least one member who moved for economic reasons (looking for a job or having a job contract), remittances were additional sources of income, constituting one element of their patchwork household economy. Money earned from migration has also allowed for greater investment in agriculture (Sobczak-Szelc and Fekih 2020).

Mobility in the case of El Faouar is indirectly motivated by the environment, as it is part of the adaptation strategy to the changes in the environment and the limitations it puts on the development of agriculture. Although there are additional sources of income available in the oasis, they are available to only a few, pushing other households to look outside the region. Therefore, El Faouar shows an example of a mix of environmental and economic, but also social-network relations that facilitate mobility.

El Ferch: from environmentally motivated to economic mobility

Settled slightly earlier than El Faouar, the oasis of El Ferch focused on vegetables and fruits rather than palm trees and cereals, with significant sheep and camel breeding. As mentioned above, the main source of water for irrigation was wells 4 m deep. In the 1970s, the oasis was connected to electricity, which allowed for the intensification of water extraction and led to the degradation of water resources, which reached its peak, and the subsequent abandonment of water extraction from traditional wells in the mid-1990s (Fig. 4).

Fig. 4
figure 4

Simplified model presenting the relation between human activity, changes in the environment and migration in El Ferch Oasis

Full size image

Like in El Faouar, the division of fields between the increasing number of descendants caused land fragmentation, decreasing income from agriculture per household and leading to the perception of agriculture as unprofitable. Facing a deterioration of living conditions and a lack of future opportunities (especially among youth), the inhabitants of El Ferch started to look for additional sources of income, mostly through mobility. The pressure to migrate was strong. Knowing examples of successful migrations from the 1960s and being tired of hard work in agriculture and all the associated challenges, inhabitants of the oasis started looking for other opportunities outside it. One of the interviewees describing the motives behind his brother’s departure frames this aptly:

He was working with me for a very long time, but he wanted a change of place. He was stressed (…) [because] the circumstances were not that good. (…) The soil is not really fertile. The land is not very large; it is about two hectares and that is insufficient. The water is very saline, at about 4.2 grams per litre. (…) Water is then insufficient to work in agriculture. (D27, M 40)

A long tradition of international migration in the region of Ghomrassen also played an additional role. Unlike Mhamid and El Faouar, in El Ferch mobility rarely affected local development, and even if migrants invested in the region, it was more in the creation of non-agricultural goods with the prospect of retirement. It was focused on building new houses in a previously agricultural area, also connected to a new water source. This affected the region in two ways. First, it further reduced the areas suitable for cultivation. Second, it increased the inequalities between the inhabitants, showing that migration and alternative sources of income give better opportunities for the future.

In El Ferch, 74% of the investigated households received income from livestock breeding in addition to plant-based agriculture. Cattle are kept outside the oasis in areas where there is less of an impact from the lack of local water resources. Apart from this, only 40% of the investigated households reported receiving any additional income outside of traditional agriculture and breeding and rarely was there more than one source. Income came mostly from trade (21% of those reporting additional sources), work in construction (17%), remittances (14%), and other sources and services (9% and 8%, respectively).

The aspirations of the younger generation are significantly shifting. Youth are less inclined to remain in agriculture, expecting higher and rapid growth in incomes. This is a sentiment echoed by an interviewee:

They [youth] are not interested in agriculture, and there are no suitable jobs for them here anyway. They need money now because it takes three to four months to harvest crops and see a profit. (X87 M 68)

Another interviewee mentioned how easier and better incomes can be made through other economic activities, such as trade. To make agriculture comparably profitable, significant investment is needed:

Of course, trade is more profitable than agriculture. We need to develop agriculture to get more revenues. For instance, if we plant potatoes this year, it could produce about 50 million dinars, but this can only happen when agriculture is very developed.

When producers’ own resources are lacking, agricultural development depends on state aid and regionally planned and managed investments. In 2004, the Agricultural Development Group’s investment led to the El Ferch water supply system, which allows for the irrigation of part of the oasis with water from Guermessa (about 15 km west of El Ferch). However, water is delivered just to the main part of the oasis, where farmland is already fragmented, so in many cases, this investment has not increased the profitability of agriculture.

New investments in the region aimed at limiting the degradation of the natural environment are misdirected and have not inhibited mobility. Additionally, the increase in aspirations towards higher living standards and a decrease in the capacity to remain in the local area strengthen people’s desire to look further afield for the fulfilment of their needs. To curb mobility, it is necessary to increase the resilience of the local population, and this is only possible by guaranteeing other sources of income outside of agriculture, particularly for the youth. El Ferch gives a final example of the complex relations between factors influencing mobility decisions. Deterioration of the environment strengthened aspirations towards mobility already created by migration from the 1960s. Further social inequalities, lack of perspective for the future of its youth, and lack of investments in local development, including possibilities to sustain agriculture, are crucial for decisions on mobility.

So similar, yet so different—the need for contextualisation and nuanced analysis: conceptual and methodological solutions

There is a general consensus that mobility is a highly contextual phenomenon that should be considered existing within a broader livelihood framework (Castles and Miller 2009). It is therefore important to examine the degree to which mobility “is contingent upon its suitability within the entire suite of available livelihood strategies” (Morrissey 2013, p. 1502). This point, illustrated through the three cases presented above, is particularly entangled with environmental factors crucial in the mobility process. It has analytical implications, indicating not only the role of environmental degradation but also the categorisation of migrants from those areas. While it is possible to assign types of environmental migrants using the typology of Renaud et al. (2011), such labelling cannot be definitive and permanent.

Contextualising environmental and non-environmental factors

The main reason for the emigration of inhabitants of Mhamid has been the deterioration of environmental conditions, which caused a decrease in household income, with no prospects for positive change in the future and a dearth of sources of income apart from agriculture. At the same time, the inhabitants of Mhamid can hardly find any other employment. Although the role of the changes in the environment is crucial here in forcing mobility, other factors should not be omitted as strengthening the need and aspirations towards migration. Although the inhabitants of El Faouar face a similar environmental situation, and the environmental resources have become insufficient for the increasing number of inhabitants, access to groundwater resources still gives possibilities to develop sustainable agriculture. Migration results from either the need for funds for new investments or a lack of job opportunities, which is the effect of no agricultural development due to insufficient water availability. This makes it an environmentally motivated migration. Finally, the inhabitants of El Ferch have also reached the maximum potential of the environment. Environmentally motivated migrations have occurred when the deterioration of the environment leads people to leave in search of other opportunities. An important additional element in the constellation of interacting causal factors is previous migration, which both increases inequalities within the society and gives opportunities for international migration through the networks of the inhabitants. Therefore, since the beginning of the twenty-first century, factors unrelated to the environment are becoming more and more important, meaning the migration motivations of most of those leaving El Ferch currently are indirectly related to the environment.

The temporal component in migrant categorisation

Categorisation of migrants is necessary to identify and localise specific groups and introduce targeted policies (McEvoy et al. 2013). However, our analysis prompts a rethink of this approach, as such categorisation often overlooks the temporal component of migration processes. The collected data show that the decisions of residents can shift depending on the importance of various factors. For instance, in El Ferch, environmental push factors have lessened over time, while factors such as “culture, education and exposure to media and other sources of images, ideas and knowledge”, which have a tremendous impact on people’s preferences and ideas about the “good life” and thus on personal life aspirations (de Haas 2021, p. 30), have increased. The arguments were also noted in El Faouar and Mhamid, where changing environmental conditions can no longer meet an inhabitant’s basic and other needs. Conclusive categorisation of migrants fails to account for the uncertainty surrounding the role of the environmental and non-environmental factors. The distinction between these factors is often blurred, making it essential to adopt a dynamic and context-specific approach in understanding and addressing migration. Moreover, the presented findings confirm the need for nuanced research in specific contexts, as the relationship between environmental and non-environmental factors is highly complex and non-linear.

The research supports the conclusions of Gray et. al (2020) that the mechanisms linking climate and migration evolve over time “and that a distinction should be made between direct and indirect linkages” (de Longueville et al. 2020; after Piguet 2022, p. 9). This study extends the gap diagnosed by the IPCC (2022) report, showing migration changes due to both climate-induced and anthropogenically-induced hydrological changes. A further limitation of the analysis is the absence of consideration of individual perceptions of these factors. A shift in the methodological approach, incorporating capabilities and aspirations, would facilitate a more integrated multidisciplinary understanding of mobility. This approach, advocated by de Haas (2021), addresses aspects not covered by most existing approaches to the environment-migration nexus (Land 2015).

Conclusions

The results of the present study are in line with the research agenda that has moved beyond linear push–pull theories within the mobility-environmental nexus, toward “a greater appreciation for context” (Zickgraf et al. 2022, p. 1084) and non-linear relations of mixed mobility factors. The starting point of the research was natural resources-related drivers. Water scarcity, however, is only one of several factors infusing mobility. Rather than oversimplifying the relationship between factors, it is suggested to embrace it in both theoretical and methodological considerations (Zickgraf et al. 2022, p. 1084).

Although the environmental conditions in the three researched oases seemed at first glance to be similar, the analysis indicated the inhabitants’ different responses to the socio-ecological transformation of their places of residence, leading to a variety of mobility strategies, such as flight in Mhamid or adaptation in El Faouar. The analysis of the empirical data shows that it is important to understand which combinations of factors actually lead to mobility and what role environmental changes play. This is especially important in arid areas, where there is staggered environmental degradation, as the effects of environmental change are often intertwined with other triggering factors (e.g. installation of electric pumps resulting in degradation of El Ferch groundwater). As a result, the exact causes underlying or impeding mobility are difficult to distinguish, which makes it challenging to implement effective policies.

Through identifying individual heterogeneous local mobility factors and how they intersect, it is possible to unpack the black box (Müller 2015) of how the complexity of the environment-mobility nexus came to create sets of conditions that generate process out-mobility. Tracing relations and changes without an a priori focus on one factor can provide an analytical framework that gives insight into the contingent and variable enrolment of the residents of the study oases into categories of environmental migrants over time. The investigated cases show the dynamics of the intertwined factors influencing mobility, one in which environmental factors dominate and in another case, economic or others.