What is a cropping system?
Evolution of field crop ecosystems
Drylands
Soils
Five premises underlie the text of the first five chapters of this Bulletin. They are:
1. Cropping systems are designed and managed to achieve human goals so they are purposeful systems.2. Like all purposeful systems, a cropping system has generic properties that are useful to assess how it functions and how effective it is. The most relevant properties here are sustainability and productivity (efficiency). Effectiveness, stability, equitability and other properties may be equally important in other contexts.
3. Soil is the major resource upon which sustainable cropping depends. Soil management is thus a key element in the management of cropping systems, both for sustainability and productivity.
4. When discussing sustainability it is helpful to define the local soil resources and to identify measurable indicators of soil productivity that are meaningful to local farmers.
5. Interpretation of local indicators of soil productivity requires some understanding of the biological mechanisms underlying soil and crop behaviour. It is better to tailor the understanding of generic mechanisms to local situations than to apply prescriptions for one cropping system to another.
The next two sections expand the first two premises. They discuss the nature of cropping systems and how best to estimate their function and effectiveness. They also describe types of cropping systems and how one type may evolve into another.
A cropping system may be defined as a community of plants which is managed by a farm unit to achieve various human goals. The latter include food, fibre and other raw materials, wealth and satisfaction. Conway (1985) encompassed these multiple goals in the phrase 'increased social value'. Farmers are part of the system. They are able to set, or modify, their own goals, so two farms with identical climates and soils may be managed with different aims to achieve a different mix of outputs.
TABLE 1
Properties of agro-ecosystems and their assessment
|
Property |
Measure |
|
Effectiveness |
Opinion, or econometric measure of satisfaction |
|
Production |
Output of system (e.g., yield of grain per household) |
|
Productivity |
Output per unit land (e.g., yield of grain per hectare) or per unit of labour, capital or any technological input |
|
Stability |
Coefficient of variation, across region or time |
|
Sustainability |
* Resource change: rate of decrease or increase in key indicators or resources,e.g., soil pH |
|
Equitability |
Income ranking of population |
IRRI (1978) gives a more detailed definition of a cropping system: "....the crop production activity of a farm. It comprises all cropping patterns grown on the farm and their interaction with farm resources, other household enterprises and the physical, biological, technological and sociological factors or environments". This is illustrated in Figure 1 in which the cropping system is one of three sub-systems likely to be present within a broader farming system.
Food security is the most basic output from a cropping system. For a subsistence household, all other outputs are secondary. For corporate agriculture, under a sophisticated cash economy, food security is established and other goals, such as profit, predominate. Thus, the first property of a cropping system is its effectiveness; that is its ability to meet the needs of the goal-setter (the farmer or corporate manager, etc.). The effectiveness of a system is a subjective attribute which should be assessed by the goal-setters and their society. It can be estimated or quantified through surveys using ranking techniques or by econometrical devices such as shadow pricing and opportunity costing (Table 1). To illustrate the importance of assessing the effectiveness of a cropping system, consider Figure 2. Here, in a survey, 270 Indonesian households were classed into six types (e.g., aiming to be self-sufficient in food, wage-earning or in commercial agriculture) and, depending on their diet and the possibility of their diet being inadequate at some times of year, they were assigned to four classes of 'food insecurity'. Households managing cropping systems for self-sufficiency occur in all categories of food insecurity. A successful household whose cropping system supports them in a 'food secure' state, will have different opinions on the effectiveness of their cropping system from a neighbouring 'food insecure' household that also aims at self sufficiency. Furthermore, it is likely that their attitudes to risk-taking and change are likely to be quite different. This is in part because of the difference in effectiveness of the two systems. One family may be open to innovation and the other perhaps caught in a spiral of conservative thinking and poverty.
Production is a most obvious output and measure of the activity of a cropping system. It can be measured as the biological or economic output from the system, for example as the grain or cash generated. It is implicitly an output from the activity of one or more management units (e.g., families). It is also a measure of the efficiency of the management of the cropping system and can be related to productivity - measured as output per unit of input (land, labour, capital, energy) (Table 1).
Conway (1985) defines other ecosystem properties as sustainability, stability and equitability (Figure 3). Of these, sustainability, the subject of this Bulletin, is the most difficult to define and estimate. FAO (1989b) states that the goal of sustainable agriculture is to 'maintain production at levels necessary to meet the increasing aspirations of an expanding world population without degrading the environment', and that sustainability 'implies concern for the generation of income, the promotion of appropriate policies, and the conservation of natural resources'. Similarly, sustainable agriculture and rural development are defined as 'the management and conservation of the natural resource base, and the orientation of technological and institutional change in such a manner as to ensure the attainment and continued satisfaction of human needs for present and future generations'.
FIGURE 2 - Four classes of food insecurity, based on diet and possible seasonal food shortage among Indonesian households (Source: Phillips, personal communication, 1989) - Notice, as expected, households which classify themselves as growing their own food have more security while landless and support-receiving households have least security. - Food Secure
FIGURE 2 - Four classes of food insecurity, based on diet and possible seasonal food shortage among Indonesian households (Source: Phillips, personal communication, 1989) - Notice, as expected, households which classify themselves as growing their own food have more security while landless and support-receiving households have least security. -Low Food Insecurity
FIGURE 2 - Four classes of food insecurity, based on diet and possible seasonal food shortage among Indonesian households (Source: Phillips, personal communication, 1989) - Notice, as expected, households which classify themselves as growing their own food have more security while landless and support-receiving households have least security. - Moderate Food Insecurity
FIGURE 2 - Four classes of food insecurity, based on diet and possible seasonal food shortage among Indonesian households (Source: Phillips, personal communication, 1989) - Notice, as expected, households which classify themselves as growing their own food have more security while landless and support-receiving households have least security. - High Food Insecurity
FIGURE 3 Properties of agro-ecosystems (Source: Conway and Barbier, 1990) - Productivity measures valued production, stability (how it varies), sustainability (how durable it is) and equitability (how it is shared). - Productivity
FIGURE 3 Properties of agro-ecosystems (Source: Conway and Barbier, 1990) - Productivity measures valued production, stability (how it varies), sustainability (how durable it is) and equitability (how it is shared). - Stability
FIGURE 3 Properties of agro-ecosystems (Source: Conway and Barbier, 1990) - Productivity measures valued production, stability (how it varies), sustainability (how durable it is) and equitability (how it is shared). - Sustainability
FIGURE 3 Properties of agro-ecosystems (Source: Conway and Barbier, 1990) - Productivity measures valued production, stability (how it varies), sustainability (how durable it is) and equitability (how it is shared). - Equitability
Conway (1985; personal communication 1991) defines sustainability as the ability of an ecosystem to maintain productivity when subjected to a major disturbing force. He is thus forced to draw a continuum between stability (in the face of minor, expected forces) and sustainability (in response to major forces). He also indicates that to assess sustainability, one needs a measure 'of the effectiveness of internal adjustments that agro-ecosystems make in response to stresses and shocks'. A more pragmatic definition of sustainability, with which most agronomists would agree, is that 'a sustainable cropping system... is one which maintains resources, such as soil and water, while providing an adequate and economic level of production, both now and for generations to come' (Hoare 1992). Similarly, a sustainable system is one in which: (i) resources are kept in balance with their use through conservation, recycling or renewal; (ii) practices preserve agricultural resources and prevent environmental damage to the farm and off-site land, water and air; (iii) production, profits and incentives retain their importance, because not only agriculture needs to be sustained, but so do farmers and society (Poincelot 1990).
Conway's definition allows for levels, or degrees, of sustainability but constrains assessment of sustainability to measurements after shock. Hoare (1992) and others (e.g., Senanayake 1991) define a sustainable system in terms easy to relate to, but often do not accommodate degrees of sustainability, which is unhelpful for management and policy development. Here, Conway's concept is accepted that it is useful to consider the possibility of various degrees of sustainability, and the widespread, but rarely developed, implication that sustainability should be measurable in either biological or economic terms (just as in measuring productivity). It follows that a biologically sustainable system may not be economically sustainable. This is witnessed by land in various parts of Europe and elsewhere that has gone in and out of cropping on several occasions during historic times depending on economic conditions.
Here, sustainability is defined consistent with FAO (1989b), as the ability of a cropping system to maintain productivity over a long term. Since this definition implies also maintenance of resources such as soil, it requires that all participants (farmers, researchers, extension specialists and policy-makers) agree upon appropriate measures by which the sustainability of particular cropping systems are assessed.
An operational schema is given in Figure 4. Generic measures of sustainability were proposed by Conway (personal communication 1991) and others, Here, forcing, inertia, elasticity and amplitude are offered as functional measures (Table 1).
TABLE 2
Indicators of key natural resources in rainfed cropping and animal production systems (Source: Hamblin 1991a)
|
Nutrient balance |
Organic matter - rate of change |
|
Erosion |
Vegetation cover - includes trees as well as stubble |
|
Productivity, yield and quality |
Water use/efficiency - i.e. actual versus potential (in some areas the potential is much less than the actual) (biomass/grain yield/US dollars), recharge (dryland salinity and nutrient leaching) |
|
Soil structure |
Infiltration |
|
pH |
Change |
|
Energy efficiency |
Input versus output of the whole agricultural system |
|
Biological factors |
Soil macro/micro flora and fauna |
|
Farm management skills |
Understanding - a good indicator would be the understanding the farmers have of their own technical system |
TABLE 3
Key measurements of sustainability in rainfed cropping and animal production systems (Source: Hamblin 1991a)
|
Indicator |
Why |
Process/measurement |
|
Water use efficiency |
Easy to relate production to rainfall; usable for both crops and pastures |
Yield (kg/mm/ha); at scales of farm, shire and region to detect trends in time and between areas |
|
Soil health |
A direct expression of sustainability of system |
Soil analysis of pH (trends), nutrient balance, direct measurement of worm counts, surveys of microflora |
|
Farm management skills |
Understanding is needed before change occurs; good level required for financial survival |
Farm records and farm surveys; units of cash flow, debt equity, whole-farm planning |
It is important that everyone involved in an assessment of the sustainability of a cropping system should agree on an inventory of resources which encapsulates its essential features. From this, the most critical attributes and, in some cases, those most easy to estimate, can be identified (Figure 4). Table 2 illustrates the type of resource inventory that might be agreed necessary to define the sustainability of a semi-arid cropping system. Defined critical resources will differ for other systems. After defining the critical resources, it is necessary to identify those indices of sustainability that are easily and relatively-cheaply measured. Some measurements emerging from consideration of Table 2 are listed in Table 3.
FIGURE 5 - Crop yield declines with continuing soil degradation and erosion. - In this study, farmers' perceptions of the rate of decline in crop yield related well to soil loss through simulated erosion (scalping or soil removal). Source: Rickson et al. 1987.
If a system is not sustainable, remedial action depends on agreement on the extent of non-sustainability and the changes in inputs to make it sustainable. Some research is encouraging here: for example, Rickson et al. (1987) find reasonable agreement between farmers' perceptions of the impact of continued soil loss on sustainability, as estimated by the extent of decline in wheat yields (Figure 5).
Stability (or variability) is the constancy of productivity in the face of small disturbances such as seasonal or year-to-year fluctuations in weather. It is a measure of the spatial- or year-to-year robustness of a farming practice and reflects not only variations in crop response to weather and soil type, but also variations in prices of inputs (such as labour and pesticides) and of the product. Last, but not least, equitability is the property that reflects how the resources and outcomes of the cropping systems are shared among the population (Table 1).
Various terms used in the sections that follow are defined in Appendix 1. Most of the definitions are taken from IRRI (1984) and ICRAF (e.g., Rocheleau et al. 1988). Cropping systems have evolved over historical time along a catena of increasing energy inputs (including mechanization), reduced biological diversity and increased risk or instability.
Boserup (1965) developed the thesis, forwarded and embellished by others, that population growth is a precondition for agricultural change. Subsistence communities in semi-arid regions depend on herding livestock (most common in semi-arid grasslands and woodlands) or shifting cultivation. In the latter, the cultivated plot is moved to new land when fertility is depleted or the plot becomes too weedy.
Boserup's thesis, backed by worldwide experience, is that increasing population, in places partly caused by immigration (e.g., settlers moving into sub-Sahelian Africa), either: (i) reduces community mobility until the villages become static and maintain permanent fields, or (ii) already semi-permanent villages are so deprived of hunting/foraging grounds that they too need to develop permanent cropping systems. As the fallow period between crops is shortened and the fields become more permanent, two strategies are available:
· Extensification and/or increased mechanization. For example, use of draught animals or tractors.· Intensification including use of shorter fallows, perhaps coupled with use of unused poorer land. This is accompanied by increased complexity of cropping, e.g., household gardens, and environmental modification, e.g., irrigation.
Okigbo (1984) comprehensively described these changes in cropping systems, and their implications for sustainability.
Extensification of cropping requires livestock for draught, or mechanization. Extensification may encourage supplementary pasture systems outside, as well as part of, the cropping systems themselves. It is also possible to develop extensive cropping without livestock. Agriculture in northeast Argentina and parts of Brazil, for example, has shifted directly from uncultivated grazing land or unused bush to extensive, mechanized cropping. Where cropping and livestock farming develop together, there are options as to their integration. In semi-arid Australia, mechanized extensification of 60 million ha of arable land has seen little (long term) change in the ratio of crop to pasture. Pasture land has increased in area since about 1930 to stabilize at about 28 million ha in 1970. Though livestock are not used for draught, this extensification has maintained a close link between crops and pasture in an integrated ley farming system. In contrast, in the semi-arid Near East, extensive cropping has been developed by private landholders without livestock. Here pasturage remains largely in the public domain with nomadic herds and common grazing land. Thus, though Australia and the Near East both have crop and livestock subsystems, in Australia they are integrated and managed by one family or corporation but in the Near East they may be separate and owned by distinct social groups.
Intensification involves biological or environmental modification. Biological modification occurs by increasing the complexity of a local cropping system by the development of multi-storied household gardens, relay cropping or mixed cropping. It is done by individual households. It increases labour inputs, diversifies crop management (though not necessarily crop diversity which may be greatest in hunter-gatherer systems), and usually increases the range of livestock (chickens, ducks, goats, fish and sometimes cattle are common in Javanese household gardens). It may reduce insect herbivores. Importantly, it creates food and income stability. Biological intensification can also be carried out on a larger, usually corporate scale. Multi-story plantation crops, such as beans under coffee, alley cropping and agroforestry, seek high outputs per unit area and stability in ways analogous to complex household gardens. Such intensification appreciably modifies the micro-environment as a secondary outcome. For example, tree legumes were planted along fence-lines in semi-arid Ethiopia in the late 1980s to provide quality forage for cattle and income from the sale of seeds. The trees provided shade and habitats for birds, which became more prized by villagers than the forage.
TABLE 4
System properties of the home garden when compared with a rice field (Source: Conway personal communication 1991)
|
|
Home garden |
Rice field |
|
Productivity |
Higher standing biomass |
Higher gross income |
|
Stability |
Year-round production ('living granary') |
Seasonal production |
|
Sustainability |
Maintenance of soil fertility |
Heavy pest and disease attack |
|
Equitability |
Home gardens in most households |
Product to land owners |
Intensification can also be carried out by deliberate modification of the environment, most obviously through irrigation. Intensification, whether by biological or environmental modification, generally gives: (i) greater diversity of land use; (ii) some intensively-used cropping land, usually in areas with favourable hydrology, fertility or social attributes (e.g., prior settlement, easy transport); and (iii) some under-used or unused land. Cropping systems in Mediterranean lands often show these features, though there under-used land sometimes reflects lack of population pressure.
It is notable that extensification of cropping systems can create relative uniformity but intensification creates diversity. Neither, however, has a predictable social or environmental outcome. Extensification, often using monocultures, can be inherently stable and sustainable, or not. The social distribution of its benefits may be egalitarian (giving either uniform prosperity, or uneconomic, socially-deprived family farms), or it may be hierarchical with economically-strong landowners and a landless class of labourers and farmhands. Intensification at the household level benefits the family directly; intensification through village cooperatives or corporations may be advantageous or create negative social and environmental outcomes. Table 4 illustrates this for a home garden and a rice field.
FAO (1987) uses the term 'drylands' to describe climates with fewer than 120 days growing season. These are divided into 'arid drylands' with less than 75 days growing season, and 'semi-arid' areas which have from 75 to 119 days growing season (Table 5).
- Seasonal changes in indices of crop growth in tropical drylands, classed by Hutchinson et al, as having climate types I (hot, seasonally wet/dry: uppermost row) or E (warm, seasonally wet/dry: middle row), and in temperate drylands (lowermost row). Growth index ( ) is a dimensionless index from zero (no growth) to 1 (maximum growth). It is the product of a temperature index (- - - -), water index ( ) and radiation or light index (.....) calculated each week of the year from standard meteorological data and using a simple soil water balance model. The temperature index is calculated for a tropical plant having an optimum temperature for growth of 28°C (shown as T = 28 in diagram) or a temperate plant with an optimum of 19 (T = 19). (Source: Hutchinson et al. 1992)
FIGURE 7 - 1.
FIGURE 7 - 2.
FIGURE 7 - 3.
FIGURE 7 - 4.
FIGURE 7 - 5.
FIGURE 7 - 6.
TABLE 5
Drylands in developing countries by region (million ha) (Source: FAO 1987)
|
|
Growth days |
|||
|
<74 |
75-119 |
Total drylands |
Total area |
|
|
Africa |
488.0 (17%) |
231.8 (8%) |
719.8 (25%) |
2878.1 |
|
Southwest Asia |
72.6 (11%) |
61.8 (9%) |
134.4 (20%) |
677.4 |
|
Southeast Asia |
47.4 (5%) |
54.9 (6%) |
102.3 (11%) |
897.6 |
|
Central America |
62.3 (23%) |
14.5 (5%) |
76.8 (28%) |
271.6 |
|
South America |
114.6 (6%) |
142.8 (8%) |
257.4 (14%) |
1770.2 |
|
East Asia |
27.7 (3%) |
70.4 (7%) |
98.1 (10%) |
954.6 |
|
Total |
812.6 (11%) |
586.2 (7%) |
1388.8 (18%) |
7449.5 |
1 Excludes South Africa.
Parenthesis show percent of total land areas for each region, e.g. Africa.
Though most semi-arid lands border the arid or desert climates of the world (Figure 6), they support a diversity of cropping (mostly with annual crops). Because of this diversity their climate should be quantified and any variations within them delineated. Such drylands have either one or two wet seasons. Bimodal rainfall regimes are characteristically found in latitudes between the wet tropics and the unimodal more temperate, semi-arid zones. Their distribution, however, is governed more by pressure systems and land masses. For example, bimodal patterns are rarely found in east Asia. In the tropics, the wet season(s) coincide with a high solar angle except for a relatively small area in east Brazil. The tropical drylands are commonly called monsoon (in Asia) or savannah (in Africa) although there is confusion over these terms. Koppen (1900) uses 'monsoon' in a restricted sense, as a climate with a dry season but able to support rainforest. In temperate areas, there is usually a single wet season that coincides with a low solar angle (winter). These wet-and-dry climates are generally called Mediterranean and, on the drier margins, semi-arid.
The tropical drylands (to use FAO's term) are categorized I and E by Hutchinson et al. (1992); and the temperate drylands as E, D and C on the basis of seasonably of available water and temperature (Figure 6 and Table 6). This delineation of climates is based on calculations of the relative constraints of soil water, temperature and solar radiation on growth of temperate and tropical crops. Each of the indices ranges from zero (no growth) to 1 (climate optimal for growth). The water, temperature and radiation indices are multiplied together to calculate a growth index, which is an estimate of the overall suitability of the location for crop growth in a particular week or month of the year.
Hutchinson's climate classification (Figure 6), unlike more traditional, geographically-oriented classifications, reflects seasonal fluctuations of environmental factors related quantitatively to crop growth. Examples relevant to the tropical drylands and climatically-similar wet-and-dry temperate areas, are shown in Figure 7.
Tropical drylands that support cropping have mean annual water indices as low as 0.29, but have a growth index (the product of the water, temperature and radiation indices) during the highest quarter (three-month period) greater than 0.3 (commonly 0.3-0.8). This is sufficient to support good growth of a single crop. Zone I coincides with parts of the Aw, Bsh and Cwa types of Koppen (1900) and Trewartha (1968) and corresponds with the 1.3, 1.4, 1.5, 1.7 and 1.9 zones of Papadakis (1975). Hutchinson's E zone, which is not as hot (Table 6) coincides with parts of Koppen's Cfa, Cwa and Bsh types and Papadakis' (1975) zones 1.7, 2, 4.2 and 5 in the tropics.
TABLE 6
Classification of climates: critical characteristics which Hutchinson et al. (1992) use to define 34 agro-climatic zones. TI19 and TI28 are the thermal indices for temperate crops (having an optimum temperature for growth of 19°C) and tropical crops (28°C), MI is water index and GI is the growth index derived from multiplying the monthly water, temperature and radiation indices
|
Group |
Annual mean TI19 |
Highest quarter GI19 |
Annual mean TI28 |
Highest quarter GI28 |
Annual mean Ml |
|
Very cold |
| ||||
|
A |
0.09 |
0.16 |
0.00 |
0.00 |
0.82 |
|
Cold |
| ||||
|
B1 |
0.27 |
0.57 |
0.04 |
0.10 |
0.90 |
|
B2 |
0.42 |
0.72 |
0.05 |
0.13 |
0.94 |
|
Cool, dry |
| ||||
|
C1 |
0.49 |
0.13 |
0.25 |
0.06 |
0.23 |
|
C2 |
0.39 |
0.28 |
0.14 |
0.11 |
0.45 |
|
C3 |
0.56 |
0.31 |
0.27 |
0.10 |
0.45 |
|
C4 |
0.49 |
0.41 |
0.27 |
0.32 |
0.32 |
|
Cool, wet |
| ||||
|
D1 |
0.38 |
0.64 |
0.17 |
0.40 |
0.59 |
|
D2 |
0.49 |
0.57 |
0.38 |
0.61 |
0.53 |
|
03 |
0.43 |
0.78 |
0.13 |
0.33 |
0.91 |
|
04 |
0.50 |
0.79 |
0.26 |
0.56 |
0.88 |
|
05 |
0.63 |
0.66 |
0.17 |
0.18 |
0.76 |
|
Warm, seasonally wet/dry |
| ||||
|
E1 |
0.81 |
0.65 |
0.35 |
0.13 |
0.55 |
|
E2 |
0.79 |
0.37 |
0.37 |
0.09 |
0.37 |
|
E3 |
0.75 |
0.46 |
0.36 |
0.30 |
0.45 |
|
E4 |
0.79 |
0.30 |
0.56 |
0.34 |
0.32 |
|
E5 |
0.97 |
0.85 |
0.51 |
0.51 |
0.57 |
|
E6 |
0.80 |
0.19 |
0.47 |
0.12 |
0.17 |
|
E7 |
0.89 |
0.69 |
0.54 |
0.62 |
0.62 |
|
E8 |
0.90 |
0.51 |
0.60 |
0.47 |
0.29 |
|
Warm, wet |
| ||||
|
F1 |
0.57 |
0.78 |
0.43 |
0.81 |
0.91 |
|
F2 |
0.74 |
0.77 |
0.66 |
0.86 |
0.89 |
|
F3 |
0.84 |
0.81 |
0.35 |
0.51 |
0.78 |
|
F4 |
0.94 |
0.85 |
0.50 |
0.66 |
0.87 |
|
F5 |
0.98 |
0.89 |
0.63 |
0.66 |
0.96 |
|
Warm to hot, very dry |
| ||||
|
G |
0.63 |
0.10 |
0.64 |
0.12 |
0.10 |
|
Hot, dry |
| ||||
|
H |
0.56 |
0.19 |
0.89 |
0.44 |
0.20 |
|
Hot, seasonally wet/dry |
| ||||
|
11 |
0.51 |
0.48 |
0.96 |
0.87 |
0.52 |
|
12 |
0.53 |
0.40 |
0.81 |
0.82 |
0.38 |
|
13 |
0.62 |
0.60 |
0.78 |
0.87 |
0.64 |
|
14 |
0.90 |
0.80 |
0.82 |
0.78 |
0.65 |
|
Hot, wet |
| ||||
|
J1 |
0.65 |
0.66 |
0.94 |
0.84 |
0.66 |
|
J2 |
0.59 |
0.62 |
0.96 |
0.88 |
0.82 |
|
J3 |
0.68 |
0.67 |
0.94 |
0.88 |
0.97 |
FIGURE 8 - Probability of occurrence of a given rainfall intensity for three soil types at Hyderabad, India (mm/week) (Source: Norman et al. 1984, adapted from Virmani 1975)
FIGURE 8 - Probability of occurrence of a given length of growing season (weeks) for three soil types at Hyderabad, India (Source: Norman et al. 1984, adapted from Virmani 1975)
The tropical drylands have close affinities with temperate semi-arid climates where annual cropping is practised. In the latter areas, the growing season exceeds the FAO limit of 119 days, and is often as much as 150-190 days because the cropping period coincides with the season of low temperature and evaporation. The C climates of Hutchinson et al. (1992) are cool and dry while their cropping systems overlap with the dry margin of their D climates, which are cool but wetter (Table 6).
Within these environments, the length of growing season and the kind of cropping system are determined not only by rainfall distribution but by soil water-holding characteristics and topographic position. The latter affects runoff to and runon from neighbouring land. The wetter regimes within the 119-day growing season allow some relay cropping on soils with favourable water-holding characteristics, but single cropping is more typical unless runon water is available. The time available for a cropping sequence at Hyderabad, India (average annual rainfall 760 mm), as affected by soil water-holding characteristics, is shown in Figure 8.
Because the drylands occupy a spectrum of water availability and cropping patterns, they are subdivided into two or more zones by Hutchinson et al. (1992) and others. For example, in the Bolivian lowlands (below 1000 m), Cochrane (1975) distinguishes four subclasses of rainfed cropping: very humid, humid, semi-humid and dry. Within these zones soil moisture is adequate for crop growth (at a 75% rainfall probability) for 8, 7, 5 and less than 5 months per year respectively. Only the dry category fits within the FAO (1987) definition of dryland. In West Africa, Kowal and Kassam (1978) distinguish seven bioclimatic regions within the savanna. The driest of these, which supports arable cropping, fits within the FAO definition of dryland. One (the southern Saharan fringe), which has no distinct wet season and no permanent cropping, is classed within the dry tropics so is outside the scope of this Bulletin. Kowal and Kassam (1978) provide climatic boundaries which, though strictly applicable only to West Africa, illustrate the gradation in wet-and-dry climates. It includes South Guinea, with a 190-240 day rainy period; North Guinea, 140-190 days; Sudan, 90-140 days; and Southern Sahel, 70-95 days.
FIGURE 9 - Broad features of soils in the tropics (Source: FAO, World Soil Resources, 1993)
Semi-arid environments do not include those dry climates in which rainfall infiltration exceeds evapotranspiration and direct evaporation for so short a period that even the most rapidly maturing crops cannot be grown without irrigation. Hutchinson et al. (1992) consider their hot dry climate (H) too dry or very marginal for rainfed crops and their very dry climate (G) unquestionably cannot support cropping. This latter category supports dry savannah (tallgrass) vegetation in the tropics, dry temperate shortgrass steppe in temperate areas, and also includes hot deserts. Dryland cropping requires a rainy season of about 55 days, which corresponds to an average annual rainfall of less than 250 mm in temperate regions and 300-350 mm in the tropics. This, depending on soil water holding capacity, will support crop growth for 65-75 days.
Figure 9 shows the global distribution of the main soil patterns. The predominant soils of the semi-arid croplands are Luvisols and Calcisols in Africa and Australia, Mollisols and Luvisols in temperate areas of the Americas, Ferralsols in wet-and-dry tropical America, Luvisols and Chernozems in temperate Eurasia and Calcisols and some Acrisols in wet-and-dry tropical Eurasia.
The climatic distribution, area and chief limitations of the main major soil orders are shown in Figure 10. All these soils occur in seasonally wet-and-dry climates, although the Acrisols are not extensive in drylands (Figure 10). Lixisols, most widespread in potential cropping areas within seasonally wet-and-dry climates, have poor water-holding capacity, hard-setting surface horizons with a risk of crusting, runoff and erosion. The topsoil texture varies from loamy sand to loam and clay, but the hard-setting properties and relatively poor infiltration and drainage (compared with, say, Acrisols) mean that crop growth is frequently limited by water availability. Figure 8 shows a delay in sowing of about two weeks on Luvisols compared with adjacent Vertisols. Lixisols often have low levels of organic matter and generally low effective cation exchange capacity (ECEC). They are commonly deficient in phosphorus.
Calcisols are extensive, particularly in north and west Africa but, lying on the dry margin of cropland, they contribute less than Luvisols to crop production in semi-arid regions. They are high in bases, in places saline, and low in organic matter. Chemozems are relatively well structured and fertile with high levels of base saturation. Their topsoil does not harden on drying, as is common in Luvisols and Lixisols. Ferralsols and Acrisols are acid, highly weathered and more common on the wetter margins than in strictly semi-arid regions. Being highly weathered, they have low fertility and ECEC though they often have high levels of exchangeable aluminium. Applied fertilizer may be leached, because of the low ECEC, or made unavailable. They have high water infiltration capacity but a relatively low water-holding capacity and are susceptible to compaction.
FIGURE 10 - Approximate climatic distribution, area and chief constraints of selected soil orders
Vertisols, dark-coloured cracking clays, though not extensive in all continents, are important in some seasonally wet-and-dry cropping systems. They have a high ECEC and base status and are often naturally fertile, though sometimes low in nitrogen, phosphorus and zinc. They have a large water-holding capacity. Their surface does not crust; they self-mulch as surface clods break down naturally on drying to small aggregates forming a good tilth. The soil cracks deeply on drying, but as it re-wets the cracks close reducing infiltration and lateral movement of water. When wet the soils are so plastic and sticky that, in Australia, Vertisols were not cropped until tractors replaced horses thus enabling them to be ploughed quickly during the short period when soil moisture is optimum.
It is useful to draw attention to the major soil types and broad climate and soil groupings, as above. For example, the extensive cereal-fallow-pasture system, covering 60 million ha of temperate semi-arid Australia, is found on four broad types of soil and differences in crop rotations largely reflect regional soil differences. However, the more common situation is that soils are linked through topography and moisture regime to landscape, as illustrated in West Africa (Figure 11).
By settling the land, cultivating and fertilizing it, humans impose other factors upon climate, soils and toposequences. Human activities, particularly clearing, inappropriate cropping and grazing, have contributed substantially to soil degradation, which is acute in much of the world's drylands (e.g. Middleton and Thomas 1992). These factors result in a large variety of cropping systems in dryland regions, which are beyond meaningful categorization. Accordingly, subsequent chapters deal with aspects of the soil resources and the mechanisms underlying cropping, so that the sustainability of various systems can be discussed in a descriptive, non-prescriptive, manner in Chapter 4.
