Impediments to Agricultural Livelihoods among Ethnic Minority Farmers in Post-War Sri Lanka

Introduction

Armed conflicts have lasting effects on the state of natural resources and the livelihoods of those who depend on them (Lamb 1985; Baumann and Kuemmerle 2016). While war-related activities may encourage or deliberately cause habitat modification and biodiversity loss (Nackoney et al. 2014; Butsic et al. 2015; Hanson 2018), the displacement that often characterizes wartime has been shown to (at least temporarily) reduce pressure on some resources (Stevens et al. 2011; Alix-Garcia et al. 2013; Sánchez-Cuervo and Aide 2013). People not only seek refuge from the immediate threats of conflict (McNeely 2003), but also to seek viable livelihoods outside of conflict zones (Raleigh 2011). In addition to the direct and indirect implications displacement has on the status of local natural resources, the degradation of infrastructure that occurs during wartime and the socio-political reorganization that armed conflict engenders may dramatically reshape future resource access and use (Benson et al. 2008; Miguel and Roland 2011; Jaafar et al. 2015).

These issues are especially pertinent to agricultural livelihoods and land uses. Displacement associated with armed conflict has been shown to cause regional-scale agricultural land abandonment (Baumann and Kuemmerle 2016) and degradation of critical infrastructure, such as irrigation systems (Dinar and Keck 1997; Jaafar et al. 2015). Yet, empirical studies elucidate heterogenous long-term impacts of war on agricultural land use, particularly after armed conflicts cease. In certain instances, such as in Iraqi Kurdistan, agricultural activity in the post-war reconstruction era parallels that of the pre-war period (Eklund et al. 2016, 2017). In Caribbean Nicaragua, extensification of farmland followed the end of the Contra War and was primarily attributed to in-migration and land colonization (Stevens et al. 2011). In contrast, only a small fraction of agricultural lands abandoned during civil conflicts in Aceh, Indonesia (Czaika and Kis-Katos 2009) or the Caucasus (Baumann et al. 2015) were re-cultivated. Similar patterns have been identified in Bosnia, where the agricultural system failed to recover in the post-conflict period (Witmer 2008)—likely, in part, because of the dangers posed by residual land mines (Andersson et al. 1995). Finally, researchers also have identified instances of ‘telecoupling’ rooted in conflict-driven land abandonment. For example, agricultural lands abandoned in South Sudan paralleled land extensification occurring near refugee camps in Northern Uganda (Gorsevski et al. 2012) and elsewhere in the region (Hagenlocher et al. 2012; Alix-Garcia et al. 2013). Similarly, spatial variation in land-use patterns driven by the activities of armed groups has been observed in Columbia (Sánchez-Cuervo and Aide 2013).

Ultimately, these dramatically varied impacts of armed conflict on agricultural land use depend on a multitude of factors, many of which are particular to the local context. However, in part because of the complexities and dangers associated with on-the-ground research, much of our understanding of the links between armed conflicts, land abandonment, and post-war land-use activities are derived from analyses of remotely sensed data (Witmer 2008; Stevens et al. 2011; Gorsevski et al. 2012; Hagenlocher et al. 2012; Qamer et al. 2012; Alix-Garcia et al. 2013; Sánchez-Cuervo and Aide 2013; Wilson and Wilson 2013; Baumann et al. 2015; Eklund et al. 2016; Yin et al. 2019). Few researchers have collected ground-level data, whether quantitative or qualitative, to complement the patterns identified through landscape-level data (Gorsevski et al. 2013; Eklund et al. 2017). Consequently, while the role of armed conflict as a driver of rapid land-use change is clear, there is a paucity of empirical research directed at the long-term effects of armed conflicts on agricultural and other natural resource-based livelihoods and regional food systems (Baumann and Kuemmerle 2016), and particularly research that seeks to document the socio-political and other processes that shape varied patterns observed globally.

Here, we present a case study from Sri Lanka, a country entrenched in civil war between the Sinhalese-controlled government and the Liberation Tigers of Tamil Eelam (LTTE) from 1983 to 2009. Small-scale agriculture, particularly rice farming, has been fundamental to the Sri Lankan economy, politics, and culture for millennia (Zubair 2005). Agriculture remains central to the livelihoods of over one-quarter of the nation’s 22 million people (Department of Census and Statistics 2019). However, the country’s civil war drove rural populations in regions of the country where the armed conflict raged (e.g., the north and east of the island) to abandon their agricultural lands and livelihoods as they fled to urban centers (Suthakar and Bui 2008). Following the war’s formal end in 2009, resettlement, infrastructure development, and other rural development projects have reinvigorated agricultural land use in war-affected regions (Kumar 2011). In some areas, analyses of remotely sensed data show that agricultural activity has exceeded pre-war levels (Ranagalage et al. 2020; Rathnayake et al. 2020). These findings are corroborated by the district-level rice production statistics as reported by the Sri Lankan government (Agriculture and Environmental Statistics Division 2022). Yet, paralleling post-war land-use research conducted elsewhere, the factors shaping and how livelihoods have been affected by this post-conflict re-agrarianization have yet to be explored.

Using household survey and focus group data collected in 2015 and 2016 in 25 communities throughout Sri Lanka’s Dry Zone, the main agricultural region, we assess whether households in formerly active combat zones have less viable agricultural livelihoods, as determined by lower yields and household assets, than households in areas less-directly affected by the quarter-century conflict. Additionally, we draw on a political ecology perspective to consider how socio-political dynamics before, during, and after a war, as well as the effects of the war on natural resources and infrastructure, have differently affected who has maintained or regained access to natural resources (Paulson et al. 2003). While the civil war in Sri Lanka was a conflict between the Sinhalese (Buddhist)-led government and Tamil (Hindu) ethnic minority population, it also inevitably affected the country’s ‘Moor’ (Muslim) and ‘Indian’, or ‘Up-Country’, Tamil (Hindu) populations. Therefore, we specifically aim to determine if and characterize how the conflict differently affected the agricultural livelihoods of households who identify with each of the country’s main ethno-religious groups.

Ethnic history (and conflict) in Sri Lanka

The contemporary Sri Lankan state has relatively recent origins, with independence in 1948 following Portuguese, Dutch, and British colonial periods. Yet, historical narratives that shaped and were shaped by the county’s decades-long civil war describe roots of the ethnic conflict going back millennia (Rajanayagam 1994). Some Tamil histories claim that their populations preceded the Sinhalese, with small numbers of merchants, artisans, and fisherfolk settling throughout the island (Nissan and Stirrat 1990). Regardless, neither the Tamil nor Sinhalese generally claim to be the island’s first inhabitants. Rather, Sinhalese ancestors came (c. 543 B.C.) with a fabled Indian prince, Vijaya, who was exiled from modern-day Bengal and settled alongside the island’s native Vedda people (an identity still maintained by a small group) (Gunawardana 1979). Under repeated pressure from waves of South Indian Tamil ‘invaders’ who crossed from the mainland, the Buddhist (converted in the third century) Sinhalese were ultimately pushed out of the islands northern and eastern regions as recently as 1000 years ago (De Silva 1987; Nissan and Stirrat 1990). Permanent Tamil settlements evolved into a kingdom centered in the city of Jaffna, which was sustained by small-scale rice cultivation and the region’s productive fisheries. The Sinhalese and Tamil populations inhabited distinct regions of the island over the centuries that followed, separated by jungle and an abandoned region of the central Dry Zone, which inform contemporary notions of discrete ethno-religious territories (Arasaratnam 1994).

Beginning at the end of the sixteenth century, European colonists sought to conquer, settle, and ultimately govern the island in its entirety, and in doing so worked to further accentuate and solidify historical ethnic boundaries and shape modern Sri Lankan history, politics, and life for all of the country’s populations. While the Portuguese and Dutch worked to supplant local religions with Christianity in their coastal colonial strongholds (in some cases quite successfully), the British acknowledged the sociocultural and political distinctions that characterized the island they called Ceylon (Manogaran 1994). Following their typical colonization protocol, the British formally pledged in 1815 to preserve Sinhalese (Buddhist) and Tamil (Hindu) religious and socio-political structures. Further, although ruling through a local Sinhalese monarchy in the Kingdom of Kandy, the British instated a racialized system of political representation in 1833 through the establishment of the Legislative Council of Ceylon that provided (marginal) political voice to the Tamil and Burgher (people of mixed European and Sri Lankan descent) communities (Wickramasinghe 2006).

British governors also encouraged the development of a plantation economy, focused on coffee and tea in the cool climate found in the Kandyan (now Central) highlands, and introduced a new ethno-political dynamic with the importation of indentured Tamil workers from mainland India. By the turn of the twentieth century, the continued immigration of these workers and their families brought the numbers of ‘Indian Tamils’ to nearly 500 000 people, accounting for around 12% of the total population (Wickramasinghe 2006). Despite a shared country of origin and Hindu religion, the immigration process and relationship with the British imperial project influenced the distinction of this group’s identity (Kanapathipillai 2009), often referred to as ‘Up-Country’ Tamils, from ‘Sri Lankan’ Tamils in the post-colonial era (Bass 2012).

Positioned as temporary foreign labor, the British colonial governs neglected to give political recognition to Up-Country Tamils. However, reformations to the Legislative Council in 1889 acknowledged the (mostly) Tamil-speaking ‘Moor’ (Muslim) ethno-religious population. Sri Lankan Moors are descendants of Arab and Persian maritime traders that began arriving on the island at least as early as the ninth century C.E. (Ali 1984). While concentrated in the eastern region of the island, Moor livelihoods, languages, and socioeconomic positions diversified as they settled in cities and towns throughout Sri Lanka (Pfaffenberger 1994). As a potential threat to the European trade monopoly, the Portuguese and Dutch placed restrictions on coastal Moors that encouraged inland migration, further distributing their population into Sinhalese-dominated regions (McGilvray 2016). As with Tamil and Sinhalese populations, the formal recognition granted by the British colonial government through their inclusion in the Legislative Council brought political unity among the increasingly diverse group, particularly in the post-colonial period.

The British use of ‘race-based’ representation was most fundamentally a recognition of the pluralistic nature of Sri Lankan society, and like the Legislative Councils established in other British colonies stemmed from liberal philosophy of the Victorian Era that advocated for ‘native’ inclusion in governance (Stultz 1972). However, despite the unequivocal diversity within each of these groups, the Legislative Council’s focus on ethno-religious categories resulted in the development of political blocks based on these specific divisions competing for power as the colonial period came to an end (Wickramasinghe 2006), which revived and strengthened the historical territorial separation between the Tamil (Hindu) and Sinhalese (Buddhist) communities (Orjuela 2005). The Tamil (and other) ethno-religious minority population(s) became increasingly marginalized by the Sinhalese majority in the post-colonial, democratic Sri Lankan state through a series of formal policies, such as legislation that declared Sinhala as the only official language (Wickramasinghe 2006), inciting political organization among and anti-government actions led by Tamil youth.

Sri Lanka’s political instability began to come to a head in 1971 with a failed insurrection attempt by a Marxist group and waivered through a series of violent riots until the onset of the civil war in 1983 (Wickramasinghe 2006). The Liberation Tigers of Tamil Eelam (LTTE) attempted to secure sovereign territory in their historical strongholds in the north and east of the island. The LTTE’s use of extraordinary violence to establish a separatist state and the Sri Lankan military’s brutal counterinsurgency efforts over the course of the 26-year war affected the life and safety of all Sri Lankas (DeVotta 2009). Thus, wartime activities resulted in the long-term internal displacement of an estimated 800 000 Sri Lankans (Kumar 2011), and left ethno-religious minority populations, such as Moors in the Tamil regions, in a particularly vulnerable position as they were caught between the warring groups (Mcgilvray 2011; Yusoff et al. 2018). This volatility and reformulation of settlement patterns undeniably affected resource use and access during the war (Korf 2004; Bohle and Fünfgeld 2007), and there are documented cases of land encroachment during the war by both Tamil and Sinhalese farmers (Korf 2003). The war ended in 2009 with the Sinhalese government declaration of victory, and domestic refugee camps were officially disbanded in 2015, releasing refugees to return home and (in theory) working to restore public utilities and rehabilitate infrastructure in war-affected communities (Kumar 2011). Yet, how post-war resettlement has affected longer-term land use and access, and whether this process has differently affected farmers who identify with each Sri Lanka’s ethno-religious groups, has remained unclear.

Materials and methods

The data used to investigate the differential, lasting impacts of Sri Lanka’s armed conflict on the agricultural livelihood of members of the nation’s various ethno-religious groups was collected as part of a broader research effort to better understand and improve agricultural decision making and adaptation to precipitation trends in Sri Lanka (ADAPT-SL). Climate change is increasingly affecting rainfall patterns in Sri Lanka, resulting in some monsoon seasons being marked by severe droughts while others experience devastating flooding events (Gunda et al. 2016). This seasonal variability poses serious challenges to small-scale farmers, who serve as the foundation of Sri Lanka’s food supply (Williams and Carrico 2017). ADAPT-SL was a collaborative effort between academic researchers and various offices and officials within the Sri Lankan national government that worked to identify and advance effective practices that both farmers and the government were implementing to combat climate variability.

Results

Discussion

While war-related displacement often temporarily relieves pressure on natural resources, research has identified dramatic variation in post-war resettlement and natural resource use patterns. This is particularly apparent among agricultural communities and livelihoods; in some instances agricultural activities fail to resume following a conflict (Witmer 2008; Czaika and Kis-Katos 2009; Baumann et al. 2015), while elsewhere agricultural land use has been observed to parallel or extend beyond pre-war levels (Stevens et al. 2011; Eklund et al. 2016, 2017). However, because most insights about how war affects natural resource and land-use patterns are derived from remotely sensed data (Witmer 2008; Stevens et al. 2011; Hagenlocher et al. 2012; Alix-Garcia et al. 2013; Baumann et al. 2015; Jaafar et al. 2015; Eklund et al. 2016; Yin et al. 2019), we lack a thorough understanding of how the socio-political dynamics of war and post-war (re-)organization may shape land and resource access. The study presented here enables us to not only detect geographic variation in agricultural livelihood viability in post-war Sri Lanka, but also recognize how the lasting effects of war are exacerbating the vulnerability of rural Sri Lanka’s ethno-religious minority populations.

Previous research utilizing remotely sensed data shows that despite land abandonment during periods of the Sri Lankan civil war (Suthakar and Bui 2008), re-agrarianization has occurred in war-affected regions with extensification above pre-war levels in some areas (Ranagalage et al. 2020; Rathnayake et al. 2020). Yet, it is clear from our analysis of household- and community-level data from throughout post-war Sri Lanka that the lingering effects of war result in variation in the viability of agricultural livelihoods. Resettlement has indeed occurred in communities directly affected by the armed conflict, but as hypothesized, farming households in these communities have lower rice yields and fewer assets than households in communities indirectly affected by the war. Further, both in general and specifically within the former conflict zone, we found that some communities have fared much better than others in the re-agrarianization process and that the observed patterns fall along ethno-religious lines.

Sinhalese (Buddhist) farming households have more viable agricultural livelihoods than Tamil (Hindu) or Moor (Muslim) households. While all of the Tamil and Moor household in our sample reside in communities that were directly affected by the war, only a small subset of Sinhalese communities are within the former conflict zone. Notably, Sinhalese households in these two communities reported higher yields than the average for both Tamil and Moor household also in the conflict zone and Sinhalese households throughout the Dry Zone. Although the virtual ethnic homogeneity of communities makes it challenging to rule out a spurious community-level effect through statistical models alone, the integration of our quantitative and qualitative results helps us to clarify and uncover some of the specific drivers of this rural ethno-religious inequality, connecting these patterns directly to socio-political phenomena related to the war. Most critical, we identified that Tamil and Moor farmers in the former conflict zone have significantly higher proportions of land that is purely rainfed than do farmers in the Sinhalese communities directly affected by the war. Access to either state- or community-managed irrigation systems are vital to both yala (dry) season rice production in Sri Lanka, as well as buffering rice production systems against wet-season droughts and floods (Lakmali et al. 2015; Williams and Carrico 2017). Further, as described by FGD participants in Community #14, the instability faced by households that lack access to functional irrigation systems pushes them away from rice farming towards alternative income generating activities. Critically, data collected through community focus groups highlight that this lack of irrigation water is not simply a relic of a pre-war history. Households in our sample’s most marginalized communities described the degradation of irrigation infrastructure during their period of war-driven displacement and inaction by the Sinhalese-dominated government in the years since their post-war resettlement as the primary challenges facing the viability of their agricultural livelihoods. Therefore, the war has perceivable, lasting impacts on the livelihoods of Tamil and Moor farming households and communities, while the agricultural livelihoods of Sinhalese farmers appear relatively unaffected.

Finally, our results also cast light on the particular plight of Moor farmers, which reported some of the lowest yields and the lowest household assets in our sample. Focus group data expose that farmers in one Moor community are not only stymied by government neglect, struggling like other ethno-religious minority farmers to maintain rice farming systems without (consistent) irrigation water, but that their minority status within a Tamil-dominated region has also directly affected their livelihoods. While these farmers were dislocated from their rural community during the war, the neighboring Tamil farmers were not and, along with Tamils that immigrated during the war, seized the opportunity to commandeer many of their residential lands. These Moor farmers now must make an arduous daily commute from the peri-urban area where they reside to cultivate their fields in their former village, compromising their ability to protect their modest rice crops against elephants or diversify their production with alternative or more drought-tolerant crops like their Tamil neighbors. The persecution of Moor communities and anti-Muslim campaigns in the post-war period, particularly in urban centers, has been noted (Ali 2011; Mcgilvray 2011, 2016; Yusoff et al. 2018). Although the number of Moors households in our study was small (reflective of the small national population), our results provide additional empirical evidence of the marginalization of these populations throughout Sri Lanka. Over the course of their long history in Sri Lanka, Sri Lankan Moors have been caught in the middle of broader political and economic conflicts and, today, continue to suffer from government inaction.

Conclusion

The findings presented here are particularly relevant to Sri Lanka, but the results have implications beyond this small island nation. Rice farming is critical to the livelihoods of people throughout rural Sri Lanka, regardless of ethno-religious identity. As witnessed in the county’s recent food crisis, driven in part by the abrupt reversal of decades-old policies that encouraged chemical fertilizer overuse, sound agricultural policies are critical to the stability of the nation’s food supply (OCHA 2022). Policymakers currently debating how to develop a more sustainable domestic food system, and economy more generally, must consider the lasting inequalities associated with the country’s multidecadal civil war on farming and farmers. This is especially important for those government entities that manage irrigation infrastructure. Water is fundamental to rice cultivation in the Dry Zone and critical to the viability of agricultural livelihoods on an island where rainfall patterns are being increasingly affected by climate change. Priority should be placed on rehabilitation projects that would increase the resilience of the region’s most marginalized communities.

Additionally, issues of post-war inequality, especially as they relate to rural recovery and redevelopment, are not unique to Sri Lanka. Yet, few studies have collected ground-level social data to understand drivers and consequences of war-related land-use change. While remotely sensed data can identify patterns, it does not provide an ability to explain important social dynamics tied to or embedded within the land-use patterns. Here, we show that the sustainability of livelihoods and food systems does not depend solely on whether land is in use, but also on who has access to that land and the resources necessary to successfully cultivate it. This work is part of a growing body of research that aims to integrate diverse methodologies and data types to understand complex natural resource problems and provide actionable insights for policymakers. As our work, which draws on a political ecology perspective, highlights, future research in this domain must explicitly consider whether and how socio-political inequalities shape and are shaped by resource access. Throughout the world, local and regional food supplies are dependent on small-scale producers (Lowder et al. 2014). Therefore, even beyond in areas recovering from periods of conflict, recognizing and working to resolve inequalities tied to rural livelihoods and food production are critical to supporting food system resilience in our era of unprecedented global environmental change.

Biographies

Author contributions

NEW: Conceptualization, Formal analysis, Data curation, Writing – Original Draft MD: Validation, Investigation, Writing – Review & Editing ARC: Methodology, Resources, Writing – Review & Editing, Supervision, Project Administration, Funding Acquisition.