Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

4. Land-use change and population in Papua New Guinea

Graham Sem

Traditional agriculture system
Land use before independence
Population dynamics
Patterns of land use


Papua New Guinea (fig. 1) comprises the eastern half of the island of New Guinea. It is a geomorphologically diverse country in the South-West Pacific Ocean and contains four large provincial islands and over 600 smaller islands. The total land area of the country is 459,854 km² (Saunders 1993) with enormous social, cultural, and biophysical diversity. The country is located on the boundary between the northward moving Australian continental plate, and the north-west moving Pacific plate, which makes it one of the tectonically active areas in the world. The main islands are characterised by block-faulted, folded, and mountainous interiors. The highest peak is Mt. Wilhelm in the Simbu province, which rises to 4,510 metres above sea level. The deltaic flood plains provide the largest areas of lowlands especially along the south coast, where freshwater swamplands are common.

Terrestrial habitats range from extensive lowlands with rainforest, savanna, grassland, and freshwater swamps to upland montane rainforests and alpine grassland (table 1). The marine and aquatic environments appear equally diverse. Papua New Guinea's native flora comprise an estimated 15,000 to 20,000 species of vascular plants, including cat 2,000 species of orchids, and more than 2,000 species of pteridophytes (Johns 1993).

Papua New Guinea culture is richly varied and people have lived in the lowlands for at least 40,000 years (Groube et al. 1986) and in the highlands for more than 24,000 years (White and O'Connell 1982). More than 750 different linguistic groups have been identified, with a variety of cultural responses to the environment. Great linguistic diversity in PNG is unmatched elsewhere and it has been suggested that the number of languages used is likely to be over 800.

Figure 1

Table 1 - Areas of Various Vegetation Types in Papua New Guinea (after Beehler 1985)

Vegetation type Area of cover (km²) % cover
Lowland rainforest 110,615 33.76
Lower montane rainforest 76,180 23.26
Savanna 22,120 6.76
Palm swamp 21,010 6.39
Herbaceous swamp 19,535 5.97
Mangrove 4,800 1.49
Montane forest 4,695 1.43
Alpine 1,720 0.53
Swamp forest 710 0.22
Strand forest 205 0.06
Total 261,590 79.87
Grassland 27,180 8.30
Gardens 21,810 6.66
Degraded forest 15,910 4.80
Plantations 1,200 0.37
Total 66,100 20.13

Traditional agriculture system

The making of history in Papua New Guinea has always been associated with its agriculture. Over 250 food-plant species have been recorded and 43 of these have been always cultivated, 51 are cultivated and harvested as wild resources, and 157 are gathered from forests, savannas, and grasslands (Paijmans 1976). Most people are subsistence agriculturalists growing mainly tuber crops and planting some fruit and nut trees. The period of cropping ranges from three to five years in the lowlands to continuous cultivation in some highland areas. The fallow period ranges from no fallow to about 25 years' fallow, although in some areas the fallow period is up to 50 years.

The majority of PNG agriculture systems are present fallow systems, or systems which have evolved from forest fallow systems. Fallow systems involve clearing and cutting forest, some burning of felled vegetation, cultivation of crops, and abandonment of the site to natural processes of regeneration.

Despite many reports to the contrary (mainly from South-East Asia) tropical forest fallow systems can be stable. Fallow systems are environmentally friendly because a tree cover over often erodible and naturally poor soils is maintained. Stable shifting agriculture systems do not destroy rainforest by cutting and burning trees; they cycle through secondary forest. Shifting cultivators avoid using previously uncultivated forests, if possible, because of the difficulties and dangers of felling the trees, and because generally secondary forest contains more useful plants and animals than primary rainforest.

Latitudinal differences in solar energy, temperature, rainfall, and soil nutrients are some of the ecological constraints on agricultural systems (Bayliss-Smith and Feachem 1977). Often a farmer is unable to directly control the constraints of solar radiation, temperature, and rainfall but is able to adapt to changes brought about by these constraints. Papua New Guinea agriculturalists have been able to adapt to changes brought about by climate change and the socio-economic conditions so that agriculture has remained the mainstay of rural societies for over 3,000 years. Some of these responses and mitigating factors have resulted in the development of most elaborate farming techniques, such as mounding, terracing, mulching, ditching, draining, and irrigation systems which are now considered to have developed independently of the major agricultural areas of the world. Some writers ascribe development of such techniques to the introduction of new crops, population growth, and increased demands for social production (sensu Brookfield 1972) and a combination of all of these factors. This technological change/innovation was a result of agricultural intensification that is now evident in Papua New Guinea.

Land use before independence

Most land-use observations before independence in 1975 were those derived from other investigations such as gold exploration and cope-plantation establishment. However, more systematic investigations of the geology and landscape, as well as land evaluation investigations were carried out by the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Division of Land Use Research. The methodology for land-evaluation investigations was based on the land systems survey (Christian and Stewart 1953). A land system is defined as a unique assemblage of the features of land, such as soils, vegetation, land-forms, rainfall, land use, and population. These surveys were designed mainly to assess the suitability of land for mechanised agriculture and pastoralism; and were based on air-photo interpretation and field checking which was limited to a representative selection of each area of the country investigated.

A total of 15 land systems were surveyed covering about 50 per cent of the total land area of the country. Land use was divided into three broad categories, subsistence cultivation, cash cropping (indigenous), and plantation. Most people in each land system were involved in subsistence cultivation. Cash cropping of mainly tree crops, such as coconuts, cocoa, coffee, and some rice, was practiced by the indigenous population in both lowland and highland areas, although in the latter some pyrethrum (Pyrethrum cinerariifolium), passion fruit, tea, and some livestock were introduced. Plantations were owned and managed by the non-indigenous population in both lowland and highland areas.

Based on land-system surveys undertaken between 1954 and 1976, it was estimated that 20 per cent of PNG land was under shifting cultivation. Some estimates of areas were given for different categories of land use in lowland and highland areas. For instance, area under subsistence cultivation per capita was estimated to be between 0.08 and 0.24 ha, with an average of 0.12 ha (McAlpine 1967, 1970). The total land area brought into commercial production (i.e. plantations) by non-indigenous people ranged from 6,400 ha in East Sepik for coconut and cocoa to over 148,000 ha for coffee and pastures in the highlands. These figures may have since changed due to the increase in population accompanied by greater economic and social demands.

Owing to the incompleteness of the population figures between 1964 and 1976, it was difficult to obtain figures for the whole country. Even if in all land-system surveys land-use intensity was not clear, it was defined for some areas, such as the highlands and the lower parts of the Sepik and Western provinces. Where sweet potato (Ipomoea batatas) was dominant in densely populated areas of the highlands, intensive, almost permanent short-fallow cultivation was evident. McAlpine (1970) suggested that the length of cultivation and fallow cycles differed greatly in the highland areas and was probably related to land pressure, environment, and cultivation techniques, although it was mentioned that no field measurements were made in support of this suggestion.

A number of major events during the pre-independence period saw an increase in the introduction of new crop varieties and emphasis on animal husbandry. The first major food and nutrition survey was conducted during 1947 and resulted in the introduction of improved pig and poultry strains and several major new projects were commenced. By 1951, the emphasis on food crops was shifted to plantation crops (McKillop 1976), but the distribution of seed of introduced vegetable crops continued.

Plantation agriculture commenced much earlier than the introduction of new food-crop varieties. The first legislation to encourage locals (Papuans) to grow cash crops was formulated in 1894 by the Lieutenant-Governor Sir William MacGregor almost 10 years after the British had annexed Papua. In 1903, indigenous cash cropping became the mainstay of the colonial government's policy concerning agricultural development. Regulations were introduced for compulsory planting of economic trees such as coconuts, rubber, and citrus trees (Waddell and Krinks 1968).

Prior to the Second World War, expatriate-owned plantations were a major source of cash-crop production, covering 24,705 ha of land. After the war, some of these plantations could not survive due to shortage of labour, low commodity prices, and high shipping costs (Crocombe 1964). Indigenous agriculture thus was to be the only hope for the future because it did not require hired labour and huge capital as was previously experienced on expatriate-owned plantations. Each village (or group of villages) was encouraged to plant cash crops as a group and, as new aspirations and increased demands for development became inevitable, the government then introduced a policy for increasing village production, which subsequently shifted away from group planting by encouraging each family to cultivate cash crops by the 1950s (Morawetz 1967; Waddell and Krinks 1968). Growth of interest in cash cropping among the people was stimulated also by the implementation of a variety of land-tenure and marketing schemes which were entirely directed towards indigenous producers. Thus, from 1951 onwards agricultural extension efforts concentrated exclusively on cash crops (Bourke et al. 1981). Land systems under various uses were therefore divided into subsistence cultivation, cash cropping, both indigenous and non-indigenous, and plantation systems which were exclusively expatriate-owned and have been described in the previous section.

As the push for cash-crop production increased so did the introduction and distribution of new food-crop varieties of sweet potato, cassava, and banana. Other crops such as peanuts, pineapples, mango, pawpaw, and guava became widely accepted in the villagers' diet (Bourke et al. 1981). During the 1970s government efforts to replace food imports were directed towards introduced vegetables, potatoes, rice and a greater emphasis on export cash-crop production. Vegetables such as tomatoes, cabbages, potato and sweet potato were cultivated on a large scale particularly for urban markets. Large-scale rice and sugar-cane production was being planned. Sugar production has been operated and managed by an overseas company since 1983.

Promoted by greater concern over increasing levels of malnutrition in rural PNG, more emphasis was placed on research and staffing, which resulted in a 1975 Nutrition Survey to assess, among other things, the levels of malnutrition in all provinces. This culminated in the formulation of the National Food and Nutrition Policy (NFNP). The main aim of the NFNP was to increase the proportion of total food supplies produced domestically. Much of the land was brought into some form of use because of the impending need for cash-crop and food production. Consequently, significant changes have occurred in food production during the last 40 years, which has led to other changes in food production and the nutrition system.

Population dynamics

It has been difficult to understand the long-term population trends of Papua New Guinea mainly because of the fact that censuses have been conducted only for a relatively short period. Censuses were conducted in 1966, 1971, 1980, and 1990. Only the latter two attempted a complete enumeration of the rural and urban populations. The population of PNG has grown from 2.2 million in 1966 to an estimated 3.8 million in 1990. This increase represents a growth rate of 2.3 per cent per annum. The estimated mid-year population for 1994 would be about 4.1 million (Hayes 1993). It appears that the population growth rate declined during the 1970s and rose slightly during the 1980s. Thus, the growth rate declined from an annual average of 2.6 per cent during the period 19661971 to 2.1 per cent in 1971-1980, and rose again slightly to 2.3 per cent between 1980 and 1990. The current officially calculated growth rate is 2.3 per cent, based on an intercensal average (Hayes 1993).

The data as presented suggest that the population growth rate has changed little over a period of 25 years, but census-based estimates of vital rates indicate that both the birth rate and death rate have declined over the same period. However, given the current population growth rate of 2.3 per cent, and the increasing threat of logging and conversion of primary rainforest, it is highly likely that more land will be brought into use by the turn of the century. The present fallow systems, which rely on low population densities and large areas of undisturbed forest, will be shortened. Assuming little or no technological change, shorter fallows will cause forest and land degradation and environmental stress. Farming of degraded lands will not be sustainable in the longer term without innovations such as introduction of new crops, new technology, and soil-fertility maintenance techniques.

The spatial distribution of the population indicates that the southern and northwestern coastal regions have low population densities (4 and 7 persons/km², respectively), while the islands and highland regions are more densely populated (10 and 22 persons km², respectively). The Western, Gulf, and West Sepik provinces remain sparsely populated. Over one-third of PNG's population is concentrated in the 13.5 per cent of the total land area in the highland region. Although 22 persons/km² is the average density for the 10 highland regions, it is reported that in some fertile highland valleys densities exceed 200 persons km² (Allen 1984), and it is in these areas that reports of "population pressure" on land have been most frequent.

Patterns of land use

Data are not available to classify land use and relate it to population density in PNG, although recent attempts have been made to classify land use by degrees of intensity (Saunders 1993). Agricultural land-use patterns vary greatly between provinces and within provinces where generalizations might misrepresent some areas (table 2).

The average cultivated area and land under use (20 per cent and 24 per cent, respectively) for the southern region is lower than the highland region (40 per cent and 43 per cent, respectively), while the north-western region has an average of 34 per cent cultivated and 42 per cent total land under use. The land area under some use in the island region is somewhat higher, with an average of 47 per cent cultivated land and 48 per cent land under some use. Thus, the total land area under any form of use is much greater than it is in the southern, north-western, and highland regions. The difference between the total cultivated land and the total used land is less than 10 per cent (table 2 and fig. 2). However, the difference in the East Sepik province exceeds this figure (17 per cent change) and this difference is mainly attributed to sago (Metroxylon sago) collection from wild stands. Sago is an important staple food (starch) in the East Sepik and other parts of lowland PNG. Again the difference in the island region is minimal, except for the West New Britain province, where there is intensive logging, raising the difference between cultivated and used land to 4 per cent. Saunders (1993) has identified 12 landuse intensity classes and divided them into six cultivated and six uncultivated land categories (fig. 3), which are described here. LU0 is defined as land with very high intensity tree crops (and others, such as sugar-cane) plantations. LU1 is also a very high intensity land use dominated by food-crop cultivation in areas of very high population density, permanent agriculture, and cultivation cycles of over five years. LU2 is a high-intensity land use where food production is the primary use in high population-density areas, and semipermanent cultivation in highland valleys and some parts of the lowland Madang and East Sepik provinces.

Table 2 - Percentage of Total Land Area: Cultivated and Used Land by Province (Source: Saunders 1993)

Province Total area (km²) Population 1990 Persons per km² 1990 Total cultivated land (%) Total used land (%) % change
Western 97,065 110,420 1 8.0 10.0 2.0
Gulf 33,847 68,122 2 11.0 12.0 1.0
Central 29,954 141,241 5 21.0 30.0 9.0
Milne Bay 14,125 158,700 11 40.0 47.0 7.0
Oro 22,510 96,239 4 19.0 22.0 3.0
Southern Highlands 25,698 317,184 12 27.0 29.0 2.0
Enga 11,839 235,561 20 31.0 37.0 7.0
Western Highlands 8,897 333,828 38 50.0 55.0 5.0
Chimbu 6,022 186,114 31 42.0 43.0 1.0
Eastern Highlands 11,006 300,515 27 50.0 53.0 3.0
Morobe 33,525 377,563 11 36.0 43.0 7.0
Madang 28,732 256,370 9 56.0 64.0 8.0
East Sepik 43,720 255,012 6 20.0 37.0 17.0
West Sepik 36,010 140,051 4 23.0 26.0 3.0
Manus 2,098 32,840 16 83.0 84.0 1.0
New Ireland 9,615 86 999 9 47.0 47.0 0.5
East New Britain 15,109 185 024 12 25.0 25.0 -
West New Britain 20,753 130,625 6 27.0 31.0 4.0
North Solomons 9,329 155,000 17 55.0 55.0 -
Total 459,854 3,567,408   25.0 30.0 5.0

Land use of moderate intensity (LU3) occurs in areas devoted primarily to food production associated with moderate population density, with short to moderately long fallow periods. Low-intensity land use (LU4) is common in low population-density areas with moderately long fallow periods in all lowland provinces and some highland areas. Very low intensity (LU5 and LU6) of land use occurs in areas of low to very low population densities, especially in lowland areas where alternative sources of food (e.g. sago and fish) supplement the cultivated crops; and where gardens are widely scattered, especially in the Western province and parts of the West Sepik province.

The uncultivated land is classified as grassland (LU7), sago stands (LU8), subalpine grassland (LU10), and savanna woodland (LU11). Grasslands have been developed and maintained as a result of burning, while sago stands provide a staple food source in large wetland areas of East Sepik, Western, and Gulf provinces. Subalpine grasslands occur in inter-montane valleys and basins of the highlands between elevations of 2,500 and 3,000 metres above sea level, while alpine grasslands occur above the tree line (above 3,400 m). Savanna woodland areas are dominated by a distinct dry-season Aw (Köppen system) climate, mainly in the southern parts of the mainland and parts of the Markham valley.


PNG is an agricultural country and it is likely that its increasing population will continue to rely heavily on agriculture in the year 2000 and beyond. More land will be brought into production and it is in the dense population areas where population pressure is becoming a problem. Despite critical shortages of land that have been reported in the highlands, no systematic treatment has been undertaken to classify land-use related to population density. However, recent data from the Agricultural Land Use Survey for PNG suggest that several areas could face similar problems of land use. Therefore, a greater research effort is required to assess and evaluate land-use change and to relate it to population growth in PNG. Such a research programme will ascertain the long-term sustainability of the current agricultural systems in the country.

Figure 2

Figure 3 Land use intensity class

At present, the Land Management Project (LMP) based at the Department of Human Geography, the Australian National University, is identifying, documenting, mapping, and describing agricultural systems for the whole country. Data on agricultural systems are currently being incorporated into a computer database linked to mapping software which, when completed for the whole country, will supplement biophysical data already contained in the Papua New Guinea Resource Information System (PNGRIS). The LMP will produce information at provincial and national levels on many aspects of subsistence agriculture, which will enable detailed assessment of the processes of intensification and the long-term sustainability of subsistence agriculture in Papua New Guinea.


Allen, B.J. (ed.). 1984. Agricultural and Nutritional Studies on the Nembi Plateau, Southern Highlands. Department of Geography, University of Papua New Guinea.

Bayliss-Smith, T. and R.G. Feachem. 1977. Subsistence and Survival: Rural Ecology in the Pacific. Academic Press, London.

Beehler, B. 1985. Conservation of New Guinea Rainforest Birds. ICBP Technical Publication 4:233-247.

Bourke, R.M., B. Carrad and P. Heywood. 1981. Papua New Guinea's Food Problems: Time for Action. Department of Primary Industry Research Bulletin 29. Port Moresby.

Brookfield, H.C. 1972. Intensification and disintensification in Pacific agriculture: A theoretical approach. Pacific Viewpoint 13:30-48.

Christian, C.S. and A. Stewart. 1953. General Report on Survey of Katherine-Darwin Region 1946. CSIRO Land Research Series Report 1. Canberra.

Crocombe, R.G. 1964. Communal Cash-cropping Among the Orokaiva. New Guinea Research Bulletin 4. Australian National University, Canberra.

Groube, L., J. Chappell, J. Muke and D. Price. 1986. A 40,000 year occupation site at Huon Peninsula, Papua New Guinea. Nature 324:453-455.

Hayes, G.R. 1993. The Demographic Situation in Papua New Guinea and Its Policy Implications: An International Perspective. Proceedings of the 19th Waigani Seminar. University of Papua New Guinea. Port Moresby.

Johns, R.J. 1993. Biodiversity and conservation of the native flora of Papua New Guinea. In: B.M. Beehler (ed.), Papua New Guinea Conservation Needs Assessment, Volume 2, 1576. Government of Papua New Guinea, Department of Environment and Conservation, Boroko, Papua New Guinea.

McAlpine, J.R. 1967. Population and land use of Bougainville and Buka islands. In: Lands of Bougainville and Buka Islands, Papua New Guinea. CSIRO Land Research Series No. 20. CSIRO. Canberra.

McAlpine, J.R. 1970. Population and land use in the Goroka-Mt Hagen Area. In: H.A. Hanntjens (ed.), Lands of the Goroka-Mt Hagen Area, Territory of Papua New Guinea. CSIRO Land Research Series No. 27. CSIRO. Canberra.

McKillop, B. 1976. A History of Agricultural Extension in Papua New Guinea. Department of Primary Industry Extension Bulletin 10. Port Moresby.

Morawetz, D. 1967. Land Tenure Conversion in the Northern District of Papua. New Guinea Research Bulletin 17. Australian National University, Canberra.

Paijmans, K. 1976. New Guinea Vegetation. Australian National University Press, Canberra.

Saunders, J. 1993. Agricultural Land Use of Papua New Guinea: Explanatory Notes to Map. PNGRIS Publication No. 1. AIDAB, Canberra.

Waddell, E. and P.A. Krinks. 1968. The Organisation of Production and Distribution Among the Orokaiva. New Guinea Research Bulletin 8. Australian National University, Canberra.

White, J.P. and J.F. O'Connell. 1982. A Prehistory of Australia, New Guinea and Sahul. Academic Press, Sydney.

Contents - Previous - Next