THE STRUCTURE OF LANDHOLDING, LAND MANAGEMENT, AND DEGRADATION



The next step in our conceptual framework highlights the linkages between
landholding structure (land tenure, farm size, fragmentation/dispersion,
and fragility) and land management practices.  As shown in figure 1, land
management practices include investments in productivity and conservation
as well as patterns of land use.  

Investment strategies include the adoption of new technologies such as
irrigation, drainage, soil conservation structures, and use of chemical
fertilizers.  They may also include the abandonment of traditional
technologies or strategies such as fallow periods and the application of
manure.  Land-use changes that result from the restructuring of landholding
are of several types.  They include fallowing practices (duration and
amount of land), cropping patterns (types of crops grown, multiple
cropping, intercropping), pastoral practices, and agroforestry.  We also
address how these changes in land management subsequently affect soil loss
and fertility depletion.


Land Tenure

Land tenure defines farmers' access to land resources.  Thus, it conditions
the decisions they make about how to use land and the kinds of investments
to make.  Researchers do not adequately under-stand the intermediate role
of tenure systems in the relationship between population pressure and land
degradation.  This is partly because conventional analysis of land
degradation has understated the complexities of land tenure, especially
during rapid population growth and the restructuring of landholding.  Also,
much of the literature is based on assumptions about economic behavior
rather than on empirical evidence.  Indeed, one well-stated perspective on
land degradation begins with the neo-classical economic assumption that
markets are the best and most efficient means for allocating and managing
natural resources.  Degradation of natural resources is thus seen as the
result of faulty markets or incentive systems.

Research often cites overuse of common property resources as the cause of
land degradation.  Hardin (1972) and Clark (1974) assert that farmers
overuse a commonly-held property to compete with 
other users.  Gradual mining of the soil eventually leads to severe
degradation as described in Hardin's classic (1968) "Tragedy of the
Commons."  One policy implication emerging from this approach to
degradation is that investments in land productivity are more likely to
occur when owners farm their own lands.

By contrast, other researchers (Guillet 1981, Bullock and Baden 1977,
Trivers 1971) suggest that farmers do not have to degrade common property. 
In many areas, strong social and cultural sanctions and a communal ethic
can induce farmers to sustain rather than degrade the land.  Moreover, many
traditional tenure systems fail to conform to the rigid categories imposed
by Western researchers.  In Fiji, for example, ownership and use rights are
very strong for more productive bottom lands but virtually non-existent for
marginal lands (Rutz, 1978).  And traditional land tenure systems in the
Amazon region, now undergoing a rapid restructuring of landholding,
continue to challenge conven-tional viewpoints on agriculture and land
degradation (Alcorn 1989).  In short, some argue that indigenous
agricultural systems are "much more sophisticated than previously assumed"
(Posey 1985: 139) and thus do not always conform to conventional models.

Boserup (1965, 1987, and 1990) contends that the tenure system will evolve
naturally from communal to individual property as a result of population
pressure and the need for agricultural intensification.  Indeed, the
evolution in land tenure is necessary, Boserup hypothesizes, before
countries can achieve significant gains in agricultural output.  Once
intensification reaches a point where land improvements are necessary, the
tenure security of private property makes it possible for farmers to get
credit to finance these improvements.  

However, as individual owners acquire land, the potential grows for
concentration of land in the hands of a few.  In turn, this leads to rental
and share arrangements between large landowners and those without
sufficient productive land.  Renters are less likely to make long-term
investments, increasing the potential for degradation.  

Researchers have seen this process operate in diverse agricultural and
ecological conditions.  For example, in her examination of tenure in
southern Honduras, Stonich (1989) found that rented lands were the most
degraded.  Rented parcels there are more likely to be on steep and degraded
slopes.  Renters lack security of tenure; most have access to parcels for
no more than three years at a time.  Thus, they have little incentive or
means to invest in costly mechanical soil conservation technologies. 
Moreover, rents are high, leaving farmers with few resources to invest in
labor or other inputs even if they want.  Renters thus rarely fertilize
fields but burn them before cultivating, a labor-saving but highly
detrimental practice. 

Migot-Adholla, et al. (1990) similarly reveal that the investment behavior
of farmers in Ghana depends on the security of land tenure.  Farmers are
considerably more likely to improve lands they own, or for which they have
long-term use rights, than lands they operate under short-term use rights. 
Improvements not only include fertilizers, mulching, and irrigation but
also investments in tree crops.  In comparison to Ghanian farmers, Kenyan
farmers report higher security of land tenure and, in turn, a greater
willingness to invest in their holdings.

In 1988, the World Bank and Rwanda's Service des Enguetes et des
Statistiques Agricoles (SESA) conducted a joint study on the effects of
land tenure on agricultural production in three regions of  Rwanda. [note 6] 
Researchers wanted to learn how tenure arrangements influence farmer
investments in their holdings and how such investments then affect crop yields.
Consistent with findings cited above from Honduras, Ghana, and Kenya,
Blarel (1989) reports that Rwandan farmers were far more likely to invest
in their own fields than in fields rented from others.

[note 6: This study was conducted as a part of the same research initiative
cited above (Migot-Adholla 1990) with reference to Ghana and Kenya.]

Alternatively, there are findings that contradict the argument that tenant
farmers invest less in improvements and prefer alternative ownership
arrangements.  Yoshinori and Hayami (1989) found one such example in a
study of tenurial arrangements among small farmers in Java.  Contrary to
previous assumptions, sharecropping, as practiced under certain conditions,
was not a deterrent to investment.  

Ervin (1982) examined studies of the relationship between tenancy and soil
conservation investment in the United States.  He, too, cautions against
automatic acceptance of the view that renters and share-croppers will have
little or no incentive to invest in soil conservation.  Ervin reports no
consistent relationship between soil conservation investments and tenancy. 
Factors, such as whether the tenant is a family member rather than a
neighbor or aspires to purchase or inherit the land, can have significant
implications for investment.  However, like those cited from Honduras,
Ghana, Kenya, and Rwanda, this study suggests that the stability of tenure,
rather than ownership, is the more important factor conditioning farmers'
decisions to invest in soil productivity.

Cook and Grut (1989) raise a further challenge to assumptions that land
ownership encourages investment in their review of agroforestry practices
in Sub-Saharan Africa.  The economic argument may seem especially
convincing for investments in agroforestry that bring in return over a
longer time period.  However, this review concludes that, in parts of rural
Africa, the tenure issue may have more to do with customary rights over
land use than with formal laws and regulations.  Cook and Grut conclude
that the evidence is not entirely clear whether individual ownership
motivates farmers to invest in agroforestry technologies for soil
conservation. 

Thus, the question is not collective versus individual ownership or even
ownership versus rental.  Rather, it is more a question of obtaining
stable, long-term use rights.  These are rights which will permit farmers
to draw benefits from their investments over the long term.  Farmers'
ability to recover investments in soil productivity do tend to be less
certain when they collectively own the land or operate it under a lease
agreement.  However, the literature shows that neither constitutes a
necessary nor a sufficient condition for low levels of investment.

Turning from investments to land-use patterns, we find that land tenure
plays an equally important role in this second dimension of land management
practices.  Land-use patterns, like investments, often reflect the
stability of use rights.  Farmers operating under long-term use rights are
more likely to plant perennial crops, produce wood, or hold the land in
long fallow.  Farmers sharing land or renting under short-term agreements
are less likely to plant for the long term.  

Again, if farmers are not assured of reaping the longer-term benefits, they
will use their holdings to maximize near-term returns.  For example, the
importance of security of tenure has emerged in  studies of indigenous
agriculture in the Amazon region.  Alcorn (1990) observes that the security
of tenure there has traditionally fostered a long-fallow agricultural
system.  Newer settlers to the region, however, have limited security of
tenure.  Thus, they have developed an extractive, short-term agricultural
system, resulting in rapid depletion of soil nutrients and increased
erosion.  Land-use controls that were important to the success of slash and
burn systems in the region have broken down.   This happened because of
development policies emphasizing short-term economic growth at the expense
of diversification and sustainability (Schmink and Wood 1987).

In the absence of focused research on the interrelationships among tenure
systems, land management, and degradation, we have presented conclusions
from several studies which treat tenure systems and land management
(investments and land use) in a broader and secondary sense.  From this
review, we conclude that tenure systems profoundly affect the ways farmers
use land and invest in farming.  We view changes occurring along a
continuum from communal to individual to rented/shared land.  An increase
in investment level often arises as land-holding evolves from communal to
individual ownership.  There is a subsequent decline as short-term use
rights become more common.  Despite the widespread historical trend, there
are numerous examples which fail to conform to this pattern.  It appears
that the "stability" of tenure, rather than ownership, may be more
important in encouraging farmers to invest in soil productivity and adopt
sustainable land-use practices.


Farm Size

Farm size can affect land management in many, though sometimes inconsistent
ways.  Large holders are often more able than small holders to maintain
traditional fallowing practices.  They also can set aside a large portion
of their holdings for non-food uses such as pasture or woodlot and other
land-use practices that help control soil loss and fertility depletion. 
Moreover, because these farmers are also comparatively wealthy, they can
invest more in inputs and improvements that will raise their long-term
productivity (Grabowski 1990).  Large holders also can endure the short-
term consequences of taking land out of production to create space for
anti-erosion technologies such as grass strips, trees, and hedge rows. 
Conversely, small farms in densely-populated regions of the world have a
relative abundance of labor to construct and maintain terraces, hedge rows,
drainage ditches, and other soil conservation measures.  And those with
small holdings often need more careful management with the related
improvements in productivity.  Their lower production level puts them
closer to the margin and at greater risk should portions of their holdings
fail to produce adequate yields.

In this context, Boserup (1965) maintains that as population density
increases, land becomes scarce and farms grow smaller.  In response, she
argues, farmers must shorten fallow periods, and increase investments in
productive technologies if they are to avoid the hardships of migration
and/or a declining standard of living.  Although Boserup uses length of
fallow as the key variable in defining the degree of intensification,
inputs such as fertilizers, irrigation, and soil conservation can
substitute for long fallow periods.

There is empirical support for Boserup's paradigm reported by Maro (1988). 
He describes several changes in investment and land use which have occurred
in Tanzania as a result of decreased farm size.  Complex networks of
irrigation channels form the basis for agricultural intensification in one
area, while farmers have terraced steep slopes in others.  

Riddell and Campbell (1986) provide further evidence from their work in the
Mandara mountain region of Cameroon.  In this region, high population
densities and small farm sizes have made the development of intensive
farming systems a necessity.  Over time, farmers have developed a complex
farming system based on soil-building strategies, integration of animal
husbandry with cultivation, and soil conservation.  

Paradoxically, as more people leave the mountains to farm on the lowlands,
problems of soil degradation have begun to emerge.  A decline in population
density from out-migration has curtailed labor available for soil
conservation and manuring activitiesÄlabor necessary for maintaining the
system's productivity.  As Riddell and Campbell (1986: 86) note:
"Traditional technology that keeps tropical soils in near-continuous
production requires dense populations to ensure adequate labor.  The
Mandara material suggests that these systems collapse as soon as population
density is reduced below some critical threshold."

Stonich (1989) concludes that large and medium holders in Honduras can
leave land in fallow for longer periods.  They are also more likely to
invest in soil conservation measures than are farmers with more limited
land resources.  And Ford (1990) reports similar findings from densely-
populated Rwanda.  The observation that smaller farms rely less on fallow
periods supports Boserup's hypothesis.  Conversely, farmers' lower
investment in soil conservation reduces their prospects for increased
production.  This highlights Boserup's failure to account for other
intermediate effects, such as variations in income and land ownership, both
of which emerge from resource scarcity.

Liverman (1990) also observes that small farmers in the state of Sonora,
Mexico are more vulnerable than large holders to the effects of drought. 
In part, this is because small holders are less likely to invest in soil
conservation and inputs such as fertilizers, seeds, and irrigation
technologies.  Lower levels of investment undoubtedly reflect the prevalent
poverty of small farmers in the area.  Conversely, in many parts of Central
and South America, it is common to find an inverse relationship between
farm size and intensity of land use (Williams, 1977).  This is particularly
true where labor inputs are the crucial factor.  Khusro (1964) has
documented the same relationship in India.

Farmers also intensify agricultural production through multiple cropping,
increasing the number of cropping cycles per year.  Boserup (1987)
describes multiple cropping as a strategy to increase yields in the face of
declining holdings.  She defines it as one of the highest degrees of
agricultural intensification. Yet, somewhat paradoxically, multiple
cropping is generally not scale neutral.  Usually larger landholders use it
as they can afford increased labor costs and the necessary inputs of
fertilizers and irrigation.  Further, when introducing multiple cropping
strategies, farmers often compensate by reducing the diversity of crops and
land uses.  Increased labor and inputs for multiple cropping may reduce
investments in lower-yielding crops that are integral to the long-term
vitality of the agroecosystem.

The effect of multiple cropping on soil degradation is not entirely clear. 
Irrigation technologies can increase production and productivity in the
short and medium run, but degraded soils can damage future production. 
Erosion and nutrient loss are common consequences of multiple cropping. 
Further, the use of inappropriate technologies to maintain yields can
devastate farmland.  Salinization often occurs with improperly designed and
managed irrigation systems.  Severe salinization can waste otherwise
productive farmland for long periods.  This problem is especially acute
when it affects small holders, who have little hope of reclaiming affected
land.  Moreover, multiple cropping usually means additional tilling and
longer periods of bare soil, vulnerable to the forces of wind and water
which cause erosion.

Intercropping is a strategy where multiple crops are grown interspersed on
the same plots.  Besides raising  yields without purchased inputs, benefits
of intercropping include soil moisture retention, erosion control, and
fewer weeds and pests.  Risk minimization is an important adaptation to
population pressure that is especially crucial in drought-prone areas. 
However, as farmers adopt higher technology strategies, they may be less
apt to pursue intercropping.  Generally, large holders who utilize
imported, modern farming practices reduce the diversity of species they
plant.  In tropical regions, such as the Amazon, monocultures are extremely
vulnerable to pests, disease, and increased leaching and erosion. 
Alternatively, for small holders operating in traditional systems where
population pressure continues to diminish holdings, intercropping is a
practical strategy that also allows farmers to maintain crop diversity.

Much of the literature examined here supports Boserup's argument.  She
contends that population growth leads to smaller farms and agricultural
intensification through changes in land use and production technologies. 
Less clear is how agricultural intensification affects land degradation. 
Boserup (1976:25) asserts that environmental deterioration occurs when a
given population increases, by natural growth or immigration, until it
exceeds the carrying capacity of the land under that system.  For example,
pastoral societies may overgraze grasslands while other groups cultivate
steep hillsides, resulting in soil erosion.  But sustained demographic
growth does not always lead to environmental degradation according to
Boserup.  She says, "The possibility exists that the population, when it
outgrows the carrying capacity of the land with the existing subsistence
technology, may change to another subsistence system with a higher carrying
capacity" (1976: 25).  However, such an assertion assumes that social
groups can readily adapt traditional subsistence practices that may have
evolved over thousands of years.

Further, implicit in Boserup's argument is the assumption that extensive
adaptation can continue indefinitely and under conditions of population
pressure never experienced in human history.  However, Boserup does not
explicitly address the ability of ecological systems to adapt to changing
human uses.  Increasingly, ecologists are concluding that, despite the
resilience of nature, agro-ecological systems have limited capacity to
adjust to rapid changes in human land use.  Traditional agricultural
systems develop over long periods and may be best suited for the
environments from which they have arisen.  The loss of land can devastate
agricultural systems that depend on crop diversity.  For example, reduced
fallow periods in slash and burn agriculture can lead to wholesale
abandonment of the agricultural system and a loss of ecological stability
(Fearnside 1985).

Here, agroecological systems represent a set of interactions between human
land uses and nature.  The kinds of adaptations Boserup describes are a
departure from the agroecological system as a critical component for the
success of farming.  Agroecosystems may, simply, be unable to adjust to the
rapid and radical adaptations that Boserup asserts are a necessary part of
coping with increased population pressure.  Consequently, we may have to
accept a measurable degree of environmental deterioration.  Boserup's
perspective on environmental degradation emphasizes declining levels of
commodity production.  As long as production increases to meet the needs of
the growing population, people perceive degradation either nonexistent or
irrelevant. 

What is not taken into account is how changes in land use and investments
may affect the potential for sustainable production.  Even with adaptations
in the farming system towards greater intensification and higher
production, degradation may still be occurring.  This fact is central to
Stocking's (1984: 9) review of soil erosion and productivity.  He remarks,
"The loss may be hidden: compensated for by additional inputs, especially
fertilizers; or covered by extra labor or bringing more land into
production; or simply tolerated as ever-declining agricultural
production...." 

In summary, some changes in investments and land use associated with
agricultural intensification are beneficial, notably those designed to
improve soil conservation.  Others, which might influence short-and medium-
run increases in production, often mask a very real decline in
productivity.  Although researchers suggest that farmers with less access
to land will have excess labor for construction and maintenance of
conservation technologies, this reasoning is not born out by empirical
study.  


Fragmentation

Both agricultural policymakers and social scientists often believe the
division of farm holdings into many, disconnected, and increasingly distant
parcels is detrimental to agricultural production.  The focus of concern is
on the high cost of moving laborers, equipment, and inputs to these many
and sometimes distant holdings.  In cases where agriculture is mechanized,
there are additional problems.  One is maneuvering large equipment in small
fields; another involves production losses stemming from a high ratio of
field edges to total area.

Conversely, there is a growing minority of researchers who have underscored
the advantages to land fragmentation.  These advantages include the
farmer's ability to exploit a greater diversity of agroecological
conditions.  This, in turn, helps sequence crops and reduces the risk of
total crop failure (Bently 1990).  Igbozurike (1970) contends that
fragmentation is actually beneficial to small farmers in West Africa simply
because agroecological diversity allows for a greater number of farmers to
survive.  This occurs although very small field sizes may limit options for
crop types and the introduction of mechanized production.

How farmers view the trade-offs undoubtedly affects land use, investment
strategies, and the process of land degradation.  Trade-offs include the
greater flexibility (control over a larger number of micro-environments)
compared to increased costs (time and labor spent traveling from one parcel
to another).   However, the research literature on fragmentation
concentrates on the effects of declining farm size (Igbozurike 1970).  It
often fails to distinguish between the two processes.  There are few
empirical studies of how fragmentation influences land use, investment
strategies, and productivity.

A study by Migot-Adholla, et al. (1990) in the Anloga region of Ghana
provides one notable exception.  There, researchers observed that farmers
are more apt to invest labor and capital in fields that are closer to their
homes, usually built up on sand bars.  Because of the location of these
fields on the sand bars, they are more prone to damage from heavy rains. 
Therefore, they require more investment in flood prevention and repair. 
Susceptibility to rain damage may be one important factor in the farmer's
decision to invest in nearby fields.  

However, this pattern of investment may also reflect the "tyranny of
space," the additional costs (time spent en route, energy required to haul
materials, etc.) in improving distant parcels.  Higher investment in nearby
parcels also reflects the higher productivity and importance of sand bar
agriculture.  A second exception is Pingali's and Binswanger's (1984) study
of the returns to investments in soil conservation.  Their findings support
the conclusion that farmers usually get higher returns from their
investments in closer locations.  However, they concede that soils in
closer fields may be more productive than those located farther from the
household compound.

Thus, despite the advantages of greater agro-environmental diversity, there
may be good reason to believe that farm fragmentation prohibits farmers
from enhancing productivity.  The greater level of investment and the
increased risk of investing in distant parcels may diminish the incentives
for certain types of conservation investments.  Farm fragmentation, as a
demographically-induced change in landholding structure is, therefore,
integral to our understanding of how population pressure can lead to land
degradation.  


Fragility

Increasing cultivation of marginal lands and their subsequent degradation
is a phenomenon common to densely-populated countries around the globe
(Gregersen, et al. 1992).  In many arid and semi-arid areas, and in most
forest ecosystems in the tropics and semi-tropics, the problem is acute
(Getahun 1991).  In the absence of sufficient off-farm opportunities, rural
populations look to the process of ecological expansionÄthe exploitation of
resources formerly outside of their immediate environments (Hawley 1950).  

Migration onto marginal lands, seen here as a significant change in the
structure of landholding, is well recognized for its impact on the
environment (Hecht 1985; Millikan 1992).  Research on the conversion of
marginal lands, and on the destruction that often follows, has focused on
two substantive issues.  The first arises from increased competition
between herders and cultivators.  As a result, pastoral systems have
changed in several environmentally-important ways.  Competition has forced
pastoralists onto drier, more fragile lands.  In addition, their
integration with cultivation systems has declined as in Rwanda (Rwamasirabo
et al. 1991).  The second is the process of deforestation.  Reduced forest
cover results primarily from the conversion of forest lands for
agricultural purposes and from increased demand for fuelwood.

The particular form of environmental degradation that results from
movements onto marginal lands is quite context-specific.  In Guatemala, for
example, it is deforestation and watershed destruction.  In Sudan,
desertification and rangeland stress have followed changes in the
management of fragile lands (Bilsborrow and DeLargy 1990).  Whatever the
case, as farmers/herders attempt to increase production in fragile areas,
the dynamics of human-environment relationships in those areas change
dramatically.  

How does this shift onto fragile lands affect farmer investments and land-
use strategies?  And what resulting problems of land degradation have
emerged?  We now address these basic questions.  
We focus on two important aspects of this demographically-induced change in
the structure of landholding: 1) expansion onto previously unexploited
lands, and 2) intensification of use on fragile holdings operated by
farmers.

In situations where population growth and land scarcity have pushed farmers
to occupy mid and upper slopes, erosion problems are particularly common. 
The characteristic lightness and thinness of these soils make them
especially prone to erosion.  These characteristics also keep yields low
and diminish returns to investments in soil conservation.  Thus, a downward
spiral of low production and low investment is easily set into motion
(Pingali and Binswanger, 1984).  It begins when these marginal lands are
taken out of their traditional uses (forest, long fallow, rangeland, etc.)
and put under more intensive cultivation.  Expansion of cultivation onto
marginal lands has resulted in degradation.  This has occurred largely
because the traditional uses of these lands, rangeland, long fallow, and
forest, are less disruptive to the soil than are seasonal or annual
cropping.  Clearing these fragile areas of trees and vegetation for
cultivation leaves the bare soils most vulnerable to accelerated wind and
water erosion.  Indeed, maintaining vegetative cover is an effective means
of controlling erosion in many environments.

Crops and other types of vegetative cover vary greatly in the degree that
they protect the soil from erosion. [note 7]  Similarly, crops differ in the
types and levels of inputs they require.  As the size of farms decreases,
options for cropping become more limited, and, when forced onto marginal lands,
choices become more limited still.  Specific slope and soil characteristics not
only constrain the choices available to farmers but also condition the effect
of cropping patterns on land degradation.  

[note 7: A well-known measure that reflects this protective quality of crops is
the C-value.  The C-value compares the soil loss ratio from land utilized with
specific tillage practices and land held in tilled continuous fallow.  For any
given field, the crop cover, canopy, and tillage practices can vary throughout
the year.  The C-value represents the average soil loss ratio resulting from
these factors over the growing season.]

Land use and crop selection is a dynamic process affected by external
structures and local conditions.  Market and policy constraints affect
farmers' decisions to grow crops or employ practices ill-suited for
environments that are new to them.  As technologies change or degradation
occurs, farmers adapt by adopting practices suitable to new conditions or
by moving into ever more fragile environments.

In Rwanda, increasing land scarcity from population growth has forced many
farmers in recent decades to depart from their traditional agricultural
system.  Historically, Rwandan farmers settled along the upper ridges of
their hillsides.  Here the soils were more fertile and cultivation was
simpler than it was farther down on steeper slopes and in marshy valleys. 
As preferred lands along upper slopes became occupied, young farmers had to
choose.  They could either cultivate smaller and less fertile plots farther
down the hillside or migrate elsewhere in search of sufficient land. 
Similarly, a recent study of non-farm strategies in Rwanda (Rwanda 1988)
shows that fallow and pasture land has been declining in recent years to
increase food production (Clay and Lewis 1990).

Farmers may have converted some of the lost fallow and pasture into
woodlot.  However, other findings suggest that households with insufficient
land have to plant ever-increasing proportions of their holdings with sweet
potatoes and other tubers (Clay and Magnani 1987; Loveridge, et al. 1988). 
These tubers have a higher caloric value than do other crops.  They also
grow relatively well in poorer soils such as those found on steeper slopes
(Gleave and White 1969).  But as annual crops, they cannot compare with the
traditional woodlot and pasture uses for these slopes in controlling soil
erosion.  In fact, studies in Africa (Lewis 1985) and in Latin America
(Ashby 1985) show that they have accelerated soil loss.

Moran (1987) examined the implications of converting fragile forest land to
cultivation in the Amazon region.  The forest canopy formerly protected the
soil, but loss of nutrients and erosion has now degraded the land.  Reasons
for degradation and exploitation of these fragile Amazonian lands vary, but
all seem to link to demographic pressure.  

Short-term intensive cultivation and large pasture tracts for cattle
(Fearnside 1985; Schmink and Wood 1987) have replaced indigenous
agriculture based on long fallow cycles.  Hecht (1985) links deforestation
in the Amazon to policies intended to encourage migration to the region. 
Millikan (1992) draws attention to increased rural unemployment and
landlessness, two symptoms of population pressure, in a study of
environmental degradation in the region.  

In nearby Ecuador, Hess (1990) describes the movement of farmers into the
fragile high altitude grasslands as a result of population growth.  Farmers
there have to cultivate steeper slopes and confine their livestock to the
upper elevations.  Erosion has increased in previously uncultivated areas
and those where livestock densities have increased markedly in recent
years.

Similarly, in the Philippines, environmental degradation has occurred from
Green Revolution technologies and from farmers moving from traditional to
more marginal areas (Western 1988).  
And in Kenya, Fury (1988) reports an increase in cultivated land in areas
previously reserved for pastoralism.  Consequently, land available to
herders has diminished in both area and quality.  Elsewhere in Africa,
Manger (1990), in a study of dryland areas of Sudan, reported competition
between farmers and herders and accompanying problems of land degradation. 
He identified expanded cultivation, commercialization of agriculture, and
increasing livestock densities as three components of demographically-
induced intensification and the main cause of the area's degradation.

Increasing land use pressure, resulting partly from population growth, is
Campbell's (1981) focus of concern in a study of marginal rangelands in
Kenya.  Land-use competition between herders and cultivators there
continues to threaten the ecological stability of these fragile lands and
contributes to desertification.  Other researchers have also identified
competition between herders and cultivators as the immediate cause of land
degradation problems in other semi-arid regions of Africa (Glantz, et al.
1987; Ibrahim 1987; Little 1987; Bassett 1988; Mwalyosi 1991).  They
commonly cite demographic pressure as the precipitating cause.  

However, the changing structure of landholding that is occurring in
pastoral areas also relates to broader processes that define the political,
social and economic context of land-use change.  We must also consider
ecological variability, especially climatic variability, in marginal areas
where land-use competition is acute.