© S. Nake et al., Hosted by EDP Sciences 2025

1 Introduction

In the late 1960s and early 1970s the north coast of West New Britain Province (WNBP), Papua New Guinea (PNG), became the site of two large oil palm projects: the first at Hoskins in 1968 and the second at Bialla in 1972 (Benjamin, 1977; Hulme, 1984; Koczberski and Curry, 2003). With its very fertile soils, suitable climate and low-density population, the Hoskins area on the north coast was recommended for oil palm development by the International Bank for Reconstruction and Development (Longayroux, 1972). The two projects included the development of agro-industrial plantation estates, centralised mills to produce palm oil, wharf facilities and large agricultural land settlement schemes (LSSs) for smallholder settlers (Fig. 1). Four additional projects with smallholder schemes were established later in Popondetta, Oro Province (1976), Milne Bay (1985), Poliamba, New Ireland Province (1991) and at Ramu, Madang Province (2003). Since then, new oil palm developments without smallholder schemes have been established in East New Britain (ENB), East and West Sepik Provinces and Central Province.

This paper draws on the authors’ research among oil palm smallholders in Papua New Guinea since the mid-1990s. Over nearly 30 years, we have conducted 11 research projects covering various aspects of smallholder oil palm livelihoods, including household oil palm and food crop production, land tenure, intra-household relations in production and income distribution, food security, and migration. Methodologically, our research has involved participatory ethnographic techniques, qualitative and quantitative household surveys (including household labour allocation surveys), field observations, and biophysical assessments of oil palm and food cropping. This paper consolidates our findings on social innovation and technology adoption, which were not the primary focus of each project. Our methodological approaches are described in the references cited for each innovation. We begin by summarising the key characteristics of three groups of smallholders before focusing on the LSS settlers at the two project sites in WNBP. Then we describe how household food security and access to land for food cropping has changed over the past three decades in response to rising land and financial pressures. We conclude with a brief discussion of the role that agro-industries can play in supporting strategies to promote food and income security amongst smallholders.

Fig. 1 Map of Papua New Guinea showing the Hoskins and Bialla study sites on the island province of West New Britain (Source: modified from CartoGIS Services, ANU College of Asia and the Pacific, Australian National University) Carte de Papouasie-Nouvelle-Guinée montrant les sites d’étude de Hoskins et Bialla dans la province insulaire de West New Britain.

2 Structure of the oil palm industry

2.1 The agro-industry sector

The oil palm industry in WNBP is comprised of four main types of stakeholders. They are the agro-industries (New Britain Palm Oil and Hargy Oil Palms), the Oil Palm Industry Corporation (OPIC) which is the extension arm of the industry, PNG Oil Palm Research Association (PNGOPRA), the research arm of the industry, and the Oil Palm Growers Associations which represent the smallholder oil palm growers in each project area. The agro-industries own and operate the processing mills, plantation estates, trucks for carting oil palm fruit to their mills, and nurseries of selected palms (all planting material for estate plantations and smallholders is selected palms).

In most project areas, smallholder fresh fruit bunches (FFB) are harvested fortnightly using family labour. There is a limited window for harvesting because oil palm fruit must be processed at the mill within three days of harvesting or palm oil quality declines. Harvested FFB and loose fruit (oil palm fruitlets dislodged from the main bunch during harvesting) are carted separately in wheelbarrows to the roadside edge of the block and stacked in separate nets for weighing and collection by agro-industry trucks for processing. Typically, the agro-industry pays men for FFB and women for collecting loose fruit. FFB and loose fruit fetch the same price, and smallholders are paid fortnightly within a few days of delivery of fruit to the mill. The price for smallholder FFB is determined by the world price of palm oil, average oil extraction rate for smallholder fruit, and average transport cost. There is an agreed price formula involving the government, the companies and the Oil Palm Industry Corporation. There is some variation in the price paid to smallholders by the different agro-industries, largely because of differing transport costs between project areas.

2.2 The smallholder sector in West New Britain Province

2.2.1 The land settlement schemes

Under the smallholder land settlement scheme (LSS) developed for Hoskins, settlers were recruited from other provinces of PNG with preference given to applicants from land-short areas of the country. Each settler family was allocated a 99-year State agricultural lease over a 6.07 ha (15 acres) block of land, of which 4 ha was planted to oil palm and the remaining 2.07 ha at the rear of the block reserved for food gardening (Koczberski et al., 2001). Leases were initially issued in the name of the male household head but more recently, they have also been issued in the names of women. The 6 ha block is the legal area of the farm over which settlers can cultivate oil palm and food crops. In total, 2373 LSS blocks, all with road access, were established at Hoskins. Groupings of approximately 130-320 blocks became subdivisions, each with a central community centre containing a primary school, health centre, agricultural extension office, fresh food market area, stores, a police post and recreational facilities. Settlers were provided with loans from the Papua New Guinea Development Bank (PNGDB) for house building, oil palm seedlings, tools, land rent and living expenses while waiting for the first harvest (Jonas, 1972; Hulme, 1984). Loans were repaid by deductions from their oil palm payments when their palms came into production.

2.2.2 Village oil palm

Following the initial development and establishment of the LSS scheme at Hoskins, attention turned to customary landowners within the mill transport catchment areas. Most Village Oil Palm (VOP) growers established 2 ha blocks of oil palm on their customary land (Koczberski et al., 2009). Like LSS settlers on their leasehold blocks, villagers were assisted with financing from the PNGDB and the agro-industry to develop their blocks (Leach and Benjamin, 1984). The VOP scheme developed slowly at Hoskins, although further expansion was stimulated in 1986 following assistance from the Asian Development Bank (Christensen, 1986) (Tab. 1). A feature of the VOP blocks in WNBP is their lower productivity compared with LSS blocks (Curry and Koczberski, 2012). Also, while LSS settlers’ legal ownership of land is limited to their 6 ha blocks, VOP growers, as customary landowners, have access to much more land beyond their 2 ha oil palm blocks.

2.2.3 Customary rights purchase blocks

In addition to VOP blocks, some customary landowners began ‘selling’ land for oil palm development to migrant settlers amongst them. These are called Customary Rights Purchase (CRP) blocks and are typically 2 ha. They are under customary land use agreements (CLUAs), a much more informal and less secure arrangement than the LSS State Leasehold blocks (Koczberski et al., 2012a) (Tab. 1).

The perceived success of the Hoskins scheme led the government to set up similar oil palm schemes at Bialla (1972) and Popondetta, Oro Province (1976). The operation and structure of the Bialla and Popondetta schemes mirrored that of Hoskins whereby the agro-industry partners developed estate plantations and processing mill, and the government opened alienated land for smallholder leasehold settlement. Settler selection for the Bialla LSS followed that of Hoskins with blocks publicly advertised and priority given to applicants from land-short areas of PNG. A VOP programme was also established at Bialla (Christensen 1986). The number of smallholder blocks has increased substantially over the last two decades (Tab. 1). The Popondetta Oil Palm scheme was initiated following a recommendation that the failed Popondetta cocoa scheme be redeveloped as an agro-industry-smallholder oil palm project (Harrison Fleming Advisory Services 1973 cited in Hulme, 1984; Grieve, 1986; Curry et al., 2019).

3 Food crop cultivation amongst land settlement schemes settlers

Food crop cultivation is a central livelihood activity, especially for women, who allocate more time to food crops than to oil palm (Curry et al., 2019). Also, most food consumed daily is from smallholders’ own vegetable gardens (Tab. 2), and vegetable gardens are a key determinant of household food and nutritional security.

LSS settlers cultivate a wide range of food crops primarily for home consumption, with surplus production sold at local markets. Common crops include bananas, yams, taro, cassava, sweet potato, beans, and green leafy vegetables. Smallholders also cultivate additional high-value crops for marketing, such as peanuts, sweet potatoes, and tobacco (Koczberski et al., 2012b; Bue, 2013). Fruits like pineapples (Ananas comosus), pawpaw (Carica papaya), watermelon (Citrullus lanatus), bananas (Musa cvs.), and betel nut (Areca catechu) are also sold at local markets (Koczberski et al., 2018).

Local marketing of garden foods provides women settlers with an additional income source beyond oil palm, making it their most important income source after oil palm (Tab. 2). Women from the LSSs make up around half of the sellers at local markets, including the town market of Kimbe (Koczberski et al., 2001; Koczberski and Curry, 2005; Ryan, 2009).

When oil palm prices are low, or during oil palm replanting when one-third of palms are out of production for 2–3 years, smallholders increase their reliance on food crops as purchases of store foods decline (Koczberski et al., 2012b). In other oil palm-producing countries, food crop cultivation has also been found to be an important buffer against fluctuating oil palm prices (Orewa, 1984; Cheyns and Rafflegeau, 2005). The income from the sale of food crops also helps families meet other financial needs, such as children’s education costs (Orewa, 2008; Nchanji et al., 2016; Masure et al., 2023).

4 Changing status of food security and smallholder responses

When the LSSs were established, a 6 ha block was considered sufficient for the needs of a single family. A lease condition was that 4 ha of oil palm would be planted within two years of settling on the block, with the remaining 2 ha of ‘reserve land’ retained for food gardens (Landell Mills Ltd., 1991). Smallholders utilised the 2 ha reserve land for food crops until the early 1990s, when they began planting an additional 2 ha of oil palm, making 6 ha of oil palm in total (Benjamin, 1977; Curry et al., 2007; Koczberski et al., 2018). Nearly all LSS blocks at Hoskins and Bialla now have 6 ha of oil palm, so a standard size block with 6 ha of oil palm would have only 0.07 ha of land available on-block for food crops. The extension of oil palm plantings into the ‘reserve land’ was in response to population growth and the need to increase income to meet the cash demands of a larger resident population. The average population per LSS block grew from a mean of 7.24 in the early 1970s (Ploeg, 1972) to 20 people by 2015 with an average of 3.75 households, placing considerable pressure on land on the block for food cropping (Koczberski et al., 2019). While an additional 2 ha of oil palm enabled cash income to rise with the growing population, land pressures increased markedly, especially during periods of low oil palm prices (e.g. increased dependence on food cropping as explained above).

The greatly reduced per capita availability of land for food cropping from population growth and from planting an additional 2 ha of oil palm stimulated innovation. LSS households demonstrated their capacity to respond to rising financial pressures and shortages of land for food cropping in several ways involving technical and social innovation, and by responding to institutional innovations created by the agro-industries. These are described next.

4.1 Innovation 1 − Technical: income diversification

Employment away from the 6 ha block can be very important when oil palm prices are low, and many young men from the LSS will seek estate plantation work. When prices rise their numbers in plantation employment fall. Many settlers have established small businesses (e.g., transport, trade stores and chicken businesses) to supplement income from oil palm and local markets, especially on densely settled blocks (Bue, 2013). The economic pressure on these blocks is reflected in the increasing numbers of non-oil palm income sources as average population per block increases. LSS blocks with one non-oil palm income source (100% local markets) have a mean block population of 11.16, while blocks with three or more non-oil palm income sources have a mean population of 15.88 (Koczberski and Curry, 2005).

4.2 Innovation 2 − Institutional and social: changing gender relations and the Lus Frut Mama scheme

Like most export crops and subsistence food production in PNG, oil palm harvesting is strongly gendered (Curry et al., 2019a). Men harvest FFB and women gather the loose fruit. Until 1996, men, as the leaseholders, were paid by the agro-industry for both FFB and loose fruit. However, with rising population pressure and declining per capita oil palm income, women’s claims on loose fruit income received lower priority in income distribution which was controlled by men. With inadequate payment of women’s labour for loose fruit collection, intra-family tensions and conflict inevitably arose around oil palm paydays (Koczberski, 2007). Consequently, most women switched their labour from oil palm to vegetable cropping over which they had greater control of the income generated (Koczberski et al., 2001).

The withdrawal of women’s labour from oil palm represented a major production loss for the agro-industry. By the late 1990s, 60–70% of loose fruit was lost through under-harvesting (Lewis, 2000). In 1996 it was recognised that under-harvesting of loose fruit was because of non-payment or underpayment of women’s labour by their husbands (Lewis, 2000; Koczberski, 2007). From 1997, women began to be paid directly by the agro-industry for loose fruit which directly addressed the payment uncertainty. By 2020, 10,000 women, representing over 80% of smallholder blocks in WNBP had their own payment card (Koczberski et al., 2022).

This institutional innovation which enabled separate payment of women’s labour in oil palm production facilitated a social innovation in which gender relations were restructured (Curry et al., 2021). The social tensions that had arisen between men and women over income distribution gave way to more harmonious and cooperative relations within the family. The new payment mechanism for women was successful because it addressed rising intra-family conflict over income distribution (Koczberski, 2007).

The new payment mechanism for women directly challenged men’s authority in oil palm production. This change was in the context of a strong male ideology concerning land access and export crops in PNG (Curry et al., 2019a). On the LSS, most leases were in the name of men, and oil palm was considered to be ‘owned’ by men with all management decisions regarding the block made by them. Despite these challenges to male authority, the new payment system was widely supported by both men and women and viewed by them as a way to resolve long-standing family conflicts over the distribution of household oil palm income. With direct payment of women for loose fruit collection, the company reinstated relatively conflict-free relations between husbands and wives and amongst the wider family. In the process, some power in oil palm production and block management was ceded to women by their husbands.

4.3 Innovation 3 − Technical: extensification of agriculture

4.3.1 Food crop cultivation in new locations

With the planting of a third, 2 ha plot of oil palm and the consequent reduction of food cropping ‘reserve land’ on LSS blocks, settlers responded by cultivating crops in new locations on- and off-block. Vegetable cropping surveys on two of the earliest LSS subdivisions established at Hoskins and Bialla revealed that while almost a quarter of the total cropping area was at the rear of the block on land remaining after planting 6 ha of oil palm, smallholders were bringing into production small patches of land previously considered unsuitable for food cropping (Bue, 2013). These included sections of the block too steep for oil palm and land around house sites. Also, some smallholders were not replanting the standard number of palm seedlings (240 palms per 2 ha replant), leaving out an edge row of palms to make more land available for food crops. Food crops cultivated on this ‘converted land’ accounted for just over 10% of the total area of smallholder vegetable gardens (Fig. 2).

Fig. 2 Area of land for food cropping per block in each land tenure category on the Land settlement scheme (Source: Koczberski et al., 2018) Superficie consacrée aux cultures vivrières par bloc dans chaque catégorie foncière sur le Land settlement scheme.

4.3.2 Purchasing a ‘Land settlement scheme’ or ‘Village oil palm’ block

A very small proportion of LSS settlers managed to acquire an additional LSS block using long-term savings from oil palm income or from the savings of a close relative in waged employment. Such opportunities are rare nowadays because there is little opportunity to accumulate savings, especially when oil palm income is shared amongst several households. The few families that purchased blocks with money saved from oil palm earnings started saving very early when their children were young. Also, the expansion of LSS schemes ceased following the release in the 1990s of some new LSS blocks at Bialla (e.g., at Soi and Kabaya subdivisions − Bue, 2013). The very limited supply of LSS blocks means that the price of LSS blocks relative to VOP blocks is very high.

4.3.3 Purchasing ‘Village oil palm’ land

With the price of LSS blocks rising beyond the reach of LSS settlers, some sought less expensive land from customary landowners to convert to 2 ha Customary Rights Purchase (CRP) blocks. These informal agreements with members of customary landowning groups in nearby villages also provide access to land for food cropping and residence (Koczberski & Curry, 2005; Koczberski et al., 2012a, 2018). Unlike the state leasehold titles of LSS blocks, tenure of CRP blocks on customary land is insecure and fraught with risk, with occupancy rights of residents frequently challenged by the customary landowners and with demands for additional payments (Curry and Koczberski, 2009).

4.3.4 Food cropping on State land

Smallholders also cultivated food crops illegally on State land surrounding the LSS (e.g. agro-industrial plantation land and state forest land), some of which was categorised as ‘buffer zone’ land in riparian habitats (Fig. 2). Two separately owned blocks with a buffer zone between them would extend their control and use of this land to the mid-point between them. The area cleared for food cropping has expanded greatly over the past 15-20 years as population and land pressures on the LSS continue to build.

4.3.5 Short-term customary land agreements

Some LSS settlers made short-term informal arrangements with neighbouring customary landowners for access to cropping land. Of the total cropping area per LSS block in 2010, 9% was located on customary land (Koczberski et al., 2012b). Typically, these agreements are for food cropping for household consumption and not for sale at local markets. Usually, the cultivators give occasional ‘gifts’ of food and/or token amounts of cash to their customary landowner hosts to maintain good relationships with them (Koczberski et al., 2009).

While it is beyond the scope of this paper to comment on the environmental aspects of certain forms of extensification, such as expansion into secondary forest or riparian zones, there are efforts to limit such impacts. The West New Britain Province oil palm industry, for instance, has obtained RSPO (Roundtable on Sustainable Palm Oil) certification, which places greater pressure on industry stakeholders to minimise these impacts.

4.4 Innovation 4 − Technical: intensification of agriculture

As per capita land area for food cropping contracted through time because of population growth and the establishment of a third, 2 ha plot of oil palm, intensification of land use also occurred (Tab. 3). In the mid-to-late 1970s, prior to a third planting of oil palm, families had sufficient land for food gardening, and a 6–9 year fallow could be maintained (Koczberski et al., 2012b, 2018). Establishing a third, 2 ha plot of oil palm, virtually eliminated the fallow period. So, intensification occurred through shorter fallows, extended cultivation periods, the increased use of fertiliser and pesticides and the substitution of slower growing traditional food crops with quick-maturing food crops more tolerant of poorer soils. No growers were identified as using fertiliser on their food gardens in 2000, but many were doing so by 2010. Similarly, pesticide use in food gardens was not observed in earlier surveys. However, by 2010 pesticides were applied in 29% of 239 surveyed gardens (Koczberski et al., 2012b). It is likely that increased pesticide use was due to shortened fallow periods, extended cultivation periods, a reduction in natural habitats of pests and diseases, and a trend towards mono-cultural production of food crops, all trends likely to be conducive to pest outbreaks. Regarding crop substitution, taro (Colocasia esculenta) and yams (Dioscorea alata and D. esculenta) gave way to bananas (Musa cvs) and cassava (Manihot esculenta) (Koczberski et al., 2018; Bue, 2013). Agricultural intensification enabled smallholders to sustain food production to meet household needs, but further innovations were required due to continued population growth.

4.5 Innovation 5 − Technical: intercropping juvenile oil palm with food crops

A relatively recent and widely adopted innovation since the planting of a third, 2 ha plot of oil palm, has been the intercropping of juvenile oil palm with food crops when shade levels are low (Fig. 3). Immature oil palm is intercropped with food crops for 3–4 crop cycles for 2–3 years, sometimes with a short fallow of a few months between planting rounds. Crops intended for sale at local markets such as peanut, cabbage (Brassica spp.), beans, cucumber and taro are planted as first crops in immature stands of oil palm. Often, sweet potato is the second crop in rotation with peanut, and in some cases, peanut is replanted as a third crop. Crops intended for sale at local markets tend to be cultivated as single species plantings while those for home consumption tend to follow the traditional pattern of mixed cropping.

As the oil palm canopy closes, the range and planting density of food crops declines, and maintenance labour such as weeding is reduced and limited to harvesting of residual crops such as banana, sugarcane and cassava which are more shade tolerant (Koczberski et al., 2018; Curry et al., 2019b). Today, the oil palm replant section provides the only substantial area on-block for food cropping. In nearby villages where there are no land shortages, food crops are rarely intercropped in newly replanted VOP blocks.

With three, 2 ha plots of oil palm and staggered replanting, settlers can grow food crops on their blocks for 6-9 years (2-3 years per 2 ha plot) over a 25-year period (Fig. 4). For 16-19 years of this 25-year period, shade levels are too high for food crops within the oil palm stand. Oil palm smallholders addressed this problem through social innovation to which we now turn.

Fig. 3 One of the authors accompanied by a local extension officer inspecting food crops in an oil palm replant area L’un des auteurs, accompagné d’un agent de développement local, inspecte les cultures vivrières dans une zone de replantation de palmiers à huile.

Fig. 4 Land available for intercropping with a staggered 2 ha replanting programme over a 25-year period on a 6 ha block. There are 6 years with 2 ha available for intercropping with food crops, 9 years of low income (4 ha in production) and 16 years with high income (6 ha in production) Surface disponible pour des cultures intercalaires avec un programme de replantation échelonnée de 2 ha sur une période de 25 ans sur un bloc de 6 ha. Il y a 6 années avec 2 ha disponibles pour des cultures intercalaires avec des cultures vivrières, 9 années de faibles revenus (4 ha en production) et 16 années de revenus élevés (6 ha en production).

4.6 Innovation 6 − Social: reciprocal food gardening rights

As the 2 ha replant sections became more important for food cropping, smallholders began accessing their neighbours’ oil palm replant areas for food cropping by the creation of reciprocal gardening rights. By 2015, 55% of the total food cropping area at Hoskins LSS was in replant areas (Koczberski et al., 2018). These reciprocal rights to cultivate food crops are modelled on access rights to cropping land practised by settlers on customary land in their home villages. In the LSS setting, these reciprocal access rights were superimposed on individual leasehold tenure, and the social networks were made up of relatives, friends, neighbours, church members and people from the same language and ethnic group. Most off-block food crop growers were from nearby blocks (Fig. 5). This social innovation dramatically expanded the supply of land for food cropping by smoothing its temporal availability. For example, by having four blocks cooperate to stagger replanting, each participating block would continuously have 0.5 ha of land available for intercropping in replant sections. This is equivalent to each block having access to an additional 12 ha of land over an oil palm cultivation cycle of 25 years (Koczberski et al., 2018).

Fig. 5 The 2 ha replant block and the home blocks of off-block food gardeners. (Source: Curry et al., 2019b) Le bloc de replantation de 2 ha et les blocs d’habitation des cultivateurs vivriers hors bloc de replantation.

4.7 Innovation 7 − Institutional and technical: the 1 ha replant

With large resident populations on the LSS blocks and heavy demands on the oil palm income, growers began to postpone replanting to beyond 25 years. This was despite the availability of interest-free credit for replanting from the agro-industries, and declining yields as palms became too tall to harvest. Growers’ reluctance to replant was because their gross income was reduced by one-third by the removal of 2 ha of oil palm during replanting while repayment of the replanting loan took 30% of their remaining income. With replanting a 2 ha plot, net income for the block fell to 47% of the level prior to replanting. This coping strategy of delayed replanting led to a high proportion of oil palm stands being well past the recommended replanting age of 25 years. It also limited income opportunities for women and older men because they lacked the strength to wield heavy harvesting poles to harvest tall palms.

From January 2016 to October 2017, we initiated a 1 ha replanting trial at Bialla in association with the agro-industry and the extension agency, OPIC. Four blocks were selected to trial the 1 ha replant with growers given the option of staggered replanting of 2 ha of over-age palms, 1 ha (120 palms) initially, with the second 1 ha replanted when the first one came into production after two to three years. The agro-industry accommodated the trial by modifying their data management and financial processes (see Curry et al., 2019b for a description of the trial). The institutional context was critical to the success of the scheme, because without the agro-industry’s cooperation the trial would not have been implemented.

As anticipated, multiple benefits were generated by the 1 ha replanting strategy for both the agro-industry and smallholders (Fig. 6). Agro-industry benefits included more timely replanting so that a higher proportion of palms were in production leading to scale economies in infrastructure use such as roads, trucks and milling capacity. Fewer stands of overage and neglected palms meant fewer refuges for pests and diseases. Furthermore, with stands of palms of various height, a larger proportion of the smallholder labour force could be involved in production; and finally, interest free company loans to smallholders would be repaid much more quickly.

Fig. 6 The advantages to growers and oil palm agro-industries of staggered replanting under the 1 ha replant strategy (Source: Curry et al., 2019b) Avantages pour les producteurs et les agro-industries du palmier à huile d’une replantation échelonnée dans le cadre de la stratégie de replantation d’un hectare.

4.7.1 Improved access to land for food gardening

The 1 ha replant option effectively doubled the time that families could cultivate food crops on their own blocks (Fig. 7; Tab. 4). With staggered replanting across two cooperating blocks, each block would have ongoing access to 0.57 ha of land for intercropping. Each replant section had a mean of 4.5 people from off-block maintaining 60% of food crop plots and occupying half of the total replant area (Curry et al., 2019b). In summary, smallholders reported the following advantages:

• the period doubled when they could intercrop on their own block, thus improving food security;

• combined with reciprocal gardening rights to the replant areas on other blocks, future access to cropping land was assured when their own block was fully planted to mature oil palm;

• the greater temporal availability of land on-block provided additional income for women through production and local marketing of food crops;

• shorter distances to food gardens benefited women as primary producers of food crops;

• reduced household vulnerability to food insecurity through improved and more secure access to cropping land (and higher incomes during replanting − see below).

Fig. 7 Land available for intercropping with a staggered 1 ha replanting programme over a 25-year period on a 6 ha block. There are 12 years with 1 ha available for food gardening, no years of low income (4 ha in production), 18 years with medium income (5 ha in production) and 7 years with high income (6 ha in production) Surface disponible pour des cultures intercalaires avec un programme de replantation échelonnée de 1 ha sur une période de 25 ans sur un bloc de 6 ha. Il y a 12 années avec 1 ha disponible pour des cultures vivrières, aucune année de faibles revenus (4 ha en production), 18 années de revenus moyens (5 ha en production) et 7 années de revenus élevés (6 ha en production).

4.7.2 Easier repayment of replanting loans, higher short and long-term incomes and reduced labour demands

With a smaller loan required to replant 1 ha of palms together with more palms in production, the 1 ha option generated higher income during replanting, and debt servicing became much easier and quicker. Instead of a ratio of 2 palms in production to repay the loan for each seedling as under the conventional 2 ha replant practice, the ratio was 5 palms in production for each seedling under the 1 ha replant option (4 mature palms in production and 1 young palm newly producing). Timelier and staggered replanting meant there were more stands of younger (shorter) palms that could be harvested by women and older men. Thus, with a doubling of the period in which younger and shorter palms are in production, the period in which they could earn an income from harvesting FFB would increase significantly. Finally, with only 1 ha of seedlings to maintain at a time, maintenance of the replant section became much easier, thereby creating a better environment for the growth and development of oil palm seedlings.

The uptake of this institutional and social innovation of the 1 ha replant option was rapid. With the support of the agro-industry and the local extension agency, the Oil Palm Industry Corporation (OPIC), the replanting strategy became the preferred replanting approach for the company.

5 Discussion and conclusions

This paper explored the various ways smallholders have responded to rising land and income pressures since the late 1960s/early 1970s, within the context of a fixed area of land − the 6 ha LSS block. LSS smallholders have displayed remarkable capacity to innovate from diversifying livelihoods to adopting potentially disruptive socio-cultural changes to agricultural practices. Sometimes change has been spontaneous and relatively easy to make (e.g. applying external inputs meant for oil palm to intercropped food crops), while at other times the changes have been challenging, requiring transformation of values or lifestyles (e.g. dietary changes and restructured gender relations) and sometimes agro-industry support (e.g. new payment schemes for smallholders) (Curry et al., 2021).

If land is available to bring into production, extensification is an easier option than agricultural intensification, in the sense that with extensification, people can continue with existing agricultural practices, and gendered hierarchies of authority and decision-making can remain intact. Intensification, on the other hand, necessitates new and higher levels of inputs, and occasionally, social innovations such as improved management practices or restructured gender relations. Amongst the settlers on the LSS in West New Britain Province, the extensification option was highly constrained because of the very restricted supply of land. With extensification being a very limited option available only to a few wealthier settlers who had accumulated savings or had access to the income of a relative working off-block, intensification became the only significant option. This was evidenced by crop substitution, extended cultivation periods, shorter fallows, and the adoption of a new payment method for women. These changes were reflected in dietary shifts and alterations in gender relations.

Technical innovations such as intercropping oil palm with food crops are straightforward with beneficial outcomes for both food crops and juvenile palms. Reciprocal exchange relationships between neighbouring blocks for access to gardening land drew on collective social memory whereby reciprocal exchange relations regarding access to customary land in settlers’ home villages were adapted to the individualised leasehold title of the land settlement schemes. This was an additional layer of tenure superimposed on the rigid leasehold system.

Some innovations, like the 1 ha replant, while dependent on institutional change to be enabled, would be attractive to smallholders whether or not they were confronted with land or financial pressures. This is simply because life and day-to-day living for both men and women was made easier with the 1 ha replant option. In contrast, the new payment mechanism for women for loose fruit collection, while also dependent on institutional innovation, challenged men’s authority in oil palm production. In the absence of land and income pressures, it is unlikely that this initiative would have been countenanced by men. However, in the context of land and financial pressures, it offered men and women a way out of the conflicted gender relationships over oil palm labour and income distribution.

Finally, while this study has highlighted the surprisingly high level of adaptive capacity of smallholder farmers facing the land limitations of the 6 ha block, it should not be interpreted as smallholder farmers having the capacity to address all the land and financial pressures that confront them. The institutional context is crucial and can have a tremendous influence on the adaptive capacity of smallholder farmers. In this case study, adjustments made by the agro-industry like the introduction of the 1 ha replant option greatly increased the temporal availability of land, while the new payment initiative for women enabled the restructuring of gender relations that led to improved livelihoods and well-being of women and their families.

Acknowledgments

We thank the Australian Centre for International Agricultural research (ACIAR) for supporting this research over an extended period. We gratefully acknowledge Hargy Oil Palms and New Britain Palm Oil for their long-term support. The Oil Palm Industry Corporation (OPIC) at Bialla, Hoskins and Popondetta, provided much support over the years. The PNG Oil Palm Research Association (OPRA) was a formal partner in this research and we thank them for their long-term input and collaboration. Smallholders were central to the research and their willing participation and cooperation is gratefully appreciated by the research team.

G. Curry is grateful to UMR Innovation, CIRAD, and the MAK’IT-FIAS Fellowship programme for supporting him during the writing of this paper. MAK’IT (Montpellier Advanced Knowledge Institute on Transitions), is co-funded by the University of Montpellier and the European Union’s Horizon 2020 Marie Skłodowska-Curie Actions (co-fund grant agreement no. 945408).

We also thank Sylvain Rafflegeau for his extensive and very helpful feedback on the paper.

Finally, we thank the referees and the editor of this journal for their very constructive feedback on the paper.

References

All Tables

All Figures

	Fig. 1 Map of Papua New Guinea showing the Hoskins and Bialla study sites on the island province of West New Britain (Source: modified from CartoGIS Services, ANU College of Asia and the Pacific, Australian National University) Carte de Papouasie-Nouvelle-Guinée montrant les sites d’étude de Hoskins et Bialla dans la province insulaire de West New Britain.
In the text

	Fig. 2 Area of land for food cropping per block in each land tenure category on the Land settlement scheme (Source: Koczberski et al., 2018) Superficie consacrée aux cultures vivrières par bloc dans chaque catégorie foncière sur le Land settlement scheme.
In the text

	Fig. 3 One of the authors accompanied by a local extension officer inspecting food crops in an oil palm replant area L’un des auteurs, accompagné d’un agent de développement local, inspecte les cultures vivrières dans une zone de replantation de palmiers à huile.
In the text

Fig. 4 Land available for intercropping with a staggered 2 ha replanting programme over a 25-year period on a 6 ha block. There are 6 years with 2 ha available for intercropping with food crops, 9 years of low income (4 ha in production) and 16 years with high income (6 ha in production) Surface disponible pour des cultures intercalaires avec un programme de replantation échelonnée de 2 ha sur une période de 25 ans sur un bloc de 6 ha. Il y a 6 années avec 2 ha disponibles pour des cultures intercalaires avec des cultures vivrières, 9 années de faibles revenus (4 ha en production) et 16 années de revenus élevés (6 ha en production).

In the text

	Fig. 5 The 2 ha replant block and the home blocks of off-block food gardeners. (Source: Curry et al., 2019b) Le bloc de replantation de 2 ha et les blocs d’habitation des cultivateurs vivriers hors bloc de replantation.
In the text

	Fig. 6 The advantages to growers and oil palm agro-industries of staggered replanting under the 1 ha replant strategy (Source: Curry et al., 2019b) Avantages pour les producteurs et les agro-industries du palmier à huile d’une replantation échelonnée dans le cadre de la stratégie de replantation d’un hectare.
In the text

Fig. 7 Land available for intercropping with a staggered 1 ha replanting programme over a 25-year period on a 6 ha block. There are 12 years with 1 ha available for food gardening, no years of low income (4 ha in production), 18 years with medium income (5 ha in production) and 7 years with high income (6 ha in production) Surface disponible pour des cultures intercalaires avec un programme de replantation échelonnée de 1 ha sur une période de 25 ans sur un bloc de 6 ha. Il y a 12 années avec 1 ha disponible pour des cultures vivrières, aucune année de faibles revenus (4 ha en production), 18 années de revenus moyens (5 ha en production) et 7 années de revenus élevés (6 ha en production).

In the text