SOIL EVALUATION

THE ROLE OF SOIL SCIENCE IN LAND EVALUATION

 

C. Dorronsoro

 

Departamento de Edafología y Química Agrícola. Facultad de Ciencias.

Campus Fuentenueva. Universidad de Granada. Spain.

 

Abstract

 

A review has been made on the concepts and the methodology of land evaluation. The results of several evaluation methods have been compared, applying them to a selection of diverse soils. Although the studies on land evaluation have been conducted on a broad diversity of characteristics—not only physical but also social, economic and political—in practice, it is frequent to limit such studies to the physical medium, given the heterogeneity of these projects. Land evaluation requires a team of multidisciplinary evaluators. The difficulty of forming these teams makes it common for such studies on land evaluation to be reduced to the analysis of the physical medium of the soil, creating a certain confusion. Therefore, we propose using the term “soil evaluation” for the assessment of the soil properties as a phase prior to land evaluation, considering soil properties in their broader sense, both the intrinsic ones (those of the soil itself: depth, texture, etc.) as well as the extrinsic ones (the soil surface: topography, climate, hydrology, vegetation, use, etc.). Soil evaluation would be similar to what today is understood as land evaluation, but excluding all the social, economic and political characteristics which would be covered under the concept of “land evaluation.”

 

Key words: soil, land, agricultural use, evaluation

 

Introduction

 

As is well known, soil is a component of the natural medium that acquires its morphology and properties after a long and slow evolution after reaching an equilibrium with environmental conditions. It is, then, a natural entity, which does not anticipate human use in its evolution. Nevertheless, ever since humans during the Neolithic shifted from hunting and gathering to farming and herding, the soil has undergone intensive exploitation. Some 10,000 years of irrational soil use by humans has transpired, with no objective beyond seeking maximum yield from every kind of soil use. As a result, the soil has reached the present day intensely degraded to the point that a great part of arable land, especially in arid and semiarid regions, is in a situation of irreversible deterioration. To stop this dramatic trend, the only solution is to institute rational soil use—that is, to use each soil in a way that best suits its characteristics and to programme its management for minimal degradation. This is precisely the final aim of land evaluation.

Land evaluation is an applied classification system that assesses the capacity of the soil for its optimal use—that is, to derive maximum benefits with minimum degradation. This can be defined, according to van Diepen et al. (1991), as “any method to explain or predict the use potential of land”.       

            Different types of soils present widely different properties, and therefore the response to each use differs. Land evolution is based on the idea that this response is a function of these properties, and, hence, knowing these, we can predict the behaviour of the soil under a given use. From the study of such properties, different degrees of suitability of the soil can be inferred for each end proposed. These degrees are reflected on maps of use capacity or suitability, on which the corresponding recommendations are made for the rational planning of soil use.

            As land evaluation is intended to offer practical results that can be plotted on territorial maps, such endeavours cannot be limited to the analysis of the physical medium of the earth, but rather must be complemented by the corresponding socio-economic studies that enable cost-benefit analyses of the profitability of the land used. Thus, land evaluation enables predictions on the biophysical and economic behaviour of land for current and potential uses.

            The above implies that land evaluation must be undertaken by an interdisciplinary team comprised of soil scientists, agronomists, ecologists, geographers, sociologists and economists, among others. Such a multifarious team is difficult to join together, given such diverse scientific training, scientific terminology and focus of each specialist. Consequently, as shown by many authors (Aandhal, 1958, Bie et al., 1973; Sys, 1985; van Diepen et al., 1991, Rossiter, 1995 and 1996), complete studies (biophysical and economic) of land evaluation have to date been scarce, the great majority being restricted to examinations of the physical medium, thus leaving the term of land evaluation quite confused.

            Given the unquestionable importance of soil study in land evaluation, we propose the term “soil evaluation” for the assessment of soil properties as a phase prior to land evaluation. This involves understanding the soil properties in their broadest sense, including both the intrinsic ones (those of the soil itself—depth, texture, etc.) as well as extrinsic ones (of the soil surface—topography, climate, hydrology, vegetation, use, etc.)

            At first, the terms “soil evaluation” and “land evaluation” were used interchangeably, but soon the term land evaluation became predominant. Soil evaluation fell into disuse. We propose the use of the term soil evaluation in its broadest sense, extending its meaning to all the characteristics that affect the soil, whether these be soil properties themselves or any related to the soil surface. Soil evaluation would be similar to what today is called land evaluation, but excluding all social, economic and political characteristics.

            Soil evaluation will be analysed below from a political, social and economic perspective to develop the land evaluation that will serve as the basis for final land-use planning.

In this sense, the separation of soil evaluation, as the basis for land evaluation, has the following objectives:

i)               To facilitate the study of soil evaluation by soil scientists, who are fully trained for soil evaluation but who find difficulty in assessing the political, social and economic aspects, such as labour costs, suitability, availability of machinery, size and situation of the parcels, costs, investment, market, infrastructure, distribution, capital, different subsidies, etc..

ii)             To avoid the confusion that the current term “land evaluation” has generated for its use to evaluate only biophysical aspects of the soils.

iii)            To provide documents based on biophysical data much more stable than the aforementioned political, social and economic aspects, with broad variability depending on the social, political and commercial decisions.

iv)            To enable easy adaptation of evaluations according to situational changes, given that soil-evaluation studies based on biophysical data are rather stable.

v)             To make soil evaluation a valid pursuit in and for itself. As indicated by Rossiter (1995 and 1996), a land-evaluation plan without previously taking specific clients into account makes little sense. Land evaluation must be directed under the request of a client, whether individual, group (societies, cooperatives) or governmental (local or national), and must be backed by the necessary guidelines that permit the plan to be put into practice. However, this situation unfortunately is far from frequent. Thus, if we abide by this ideal, studies on land evaluation would be restricted to a limited number of occasions. However, we believe that soil-evaluation studies can be valid by themselves, without the need for prior request, as the soil evaluation can simply be undertaken at any moment as an important contribution to the knowledge of soil quality in a region, as an environmental resource, as important as knowledge of basic soil types, lithology, geology, hydrology, etc..

Land evaluation will be necessary both for the for a study of the regional planning of a territory as well as for the particular case of a change in the use of a specific parcel. Land evaluation must be capable of predicting the particular behaviour of a soil for new use.

However, as indicated by Rossiter (1996), to focus a land-evaluation plan, we must bear in mind that, when well formulated, the plan should provide major benefits—but, when poorly calculated, it inflicts heavy losses and, even worse, irreversible degradation. A shift in land use might could detonate “chemical time-bomb” (van Latestejin, 1998).

 

Evaluation characteristics

 

Soil evaluation can be performed directly or indirectly. In the former case, the evaluation stems from field tests (experimental stations, random agricultural sampling in the field) or from yield production provided by individual farmers (farm records) and cooperatives, or else from agricultural statistics. These data are usually local, spotty, sometimes not reliable and are generally difficult to extrapolate. Therefore, the evaluation is normally conducted indirectly on the basis of the soil properties, assuming that yield of a given soil depends on its properties and its level of management. The evaluations made in this manner should be validated finally with real yield data.

            In an indirect evaluation, it is evident that to define a degree of suitability, it does not suffice to choose only one property, but rather it requires a group of properties, possibly the more the better. The properties to choose will depend on the proposed use of the soil. The values of these evaluation parameters can be derived from disparate sources (remote sensing, maps, the literature and directly from the field or laboratory) and with sharply differing degrees of precision.

In addition, these evaluation characteristics can be combined in many ways in various assessment systems used in soil evaluation, so that even for the same use, the results for a given soil can differ markedly, depending on the evaluation system chosen. Therefore, we propose that it is useful to define in a general manner the properties that most influence soils as well as their degree of suitability. For the case of agricultural use, which is the most frequent, Table 1 highlights some of these, although other properties can be used, such as: genetic profile, parent material, clay mineralogy, microrelief, altitude, annual rainfall distribution, evapotranspiration, vegetation, etc..

 

Evaluation systems

 

A plethora of evaluation methods are used under diverse philosophies and techniques. Some methods value the degree of suitability of the properties, while others place more emphasis on the possible limiting factors for soil use. These latter methods are more consistent, since, as we know, the true limits on soil use are the negative characteristics (according to Liebig’s law of the minimum), regardless of the degree of suitability of the most favourable properties.

Some methods use qualitative criteria while others use quantitative. The qualitative systems are normally used in examinations aimed at a general evaluation of broad zones. The quantitative methods are used more often in detailed studies and thus need more information on the soils, both to construct the evaluation system as well as to apply it, but they are more objective and the results are therefore more reliable. Other methods begin with qualitative data which are weighted to reach a final numerical result.

           

 

	Very favourable	Favourable	Unfavourable	Very unfavourable
Intrinsic properties
Effective depth, cm	>120	120-70	70-30	<30
Texture	balanced	moderate heavy	heavy	light
Course fragments, %	<10	10-30	30-60	>60
Structure	f, md, 3, 2	c, 1	sg, 0	ms, 0
Compact, cemen, gr, cm	Absent	md, >60	md, >20 or st, >60	st, < 30
Available water, mm	>100	100-60	60-20	<20
Internal drainage	Without hydro	Hydro > 80 cm	Hydro > 40 cm	Hydro a 0 cm
Permeability, cm/hour	>2	2-0.5	0.5-0.1	<0.1
Organic matter, %	>5	5-2	2-1	<1
CEC, cmol(+)kg-1	>40	40-20	20-10	<10
Saturation degree, %	>75	75-50	50-25	<25
pH	7.3-6.7	6.7-5.5 or 7.3-8.0	5.5-4.5 or 8.0-9.0	<4.5 or >9.0
Carbonates, %	<7	7-15	15-25	>25
Salinity, dSm-1	<2	2-6	6-12	>12
Extrinsic properties
Slope, %	<4	4-10	10-25	>25
Surface stoniness,%	<2	2-20	2-20	>50
Surface rockiness, %	<2	2-20	2-20	>50
Flooding, months	0	<1	1-3	>3
Erosion, Tm/ha/year	<10	10-20	20-60	>60
Ploughing	no problems	limited	severe	very severe
Precipitation, mm	>1000	1000-600	600-300	<300
Frost, Tª<0º, months	<1	1-3	3-6	>6

 

Table 1. Environmental indicators of the degree of suitability of soils for agricultural use.

Texture: balanced = loam, silt loam, sandy clay loam; heavy = sandy clay, clay loam, silty clay loam, silt; heavy = clay, silty clay; light = sand, loamy sand. Structure. f = fine; md = medium; c = coarse; sg = single grain; ms = massive; 3 = strong; 2 = moderate; 1 = weak; 0 = structureless. Compact = compaction, cemen = cementation, gr = degree, cm = depth at which it appears; md = moderate; st = strong. Internal drainage: hydro = hydromorphy. CEC = cation-exchange capacity. Ploughing: no problems = ploughing is possible at any time of the year; limited = not possible during wet periods, clayey soils; severe = only in dry periods, soils very clayey. Very severe = not possible due to steep slopes or high groundwater table; Precipitation = Annual precipitation.

           

            Some systems group the classes into a series of levels of importance (order, class, subclass, type, etc.), and are thus hierarchical systems. Other systems have one category, and these are frequently parametric. In these latter systems, mathematical formulae are applied so that the final result is expressed in numerical terms. These can be additive (Index = A + B + C + D + ...) or with a multiplicative scheme ( Index = A x B x C x D x...), the latter offering better results for following the minimum law. The additive systems give an evaluation that is usually correct from the theoretic standpoint but can give evaluations that are not realistic inasmuch as they do not represent the serious consequences implicit in a highly limiting soil factor. It is generally accepted that the parametric methods are, according to McRae and Burnham (1981): simple, objective, quantitative, reliable, easy to understand and apply, even by the non-specialist, and easy to modify and adapt to new uses. Their main disadvantage is precisely that their objectivity and precision are illusory. Their formulation is difficult and, if they are not well applied, their results can be completely erroneous. The scheme is too artificial and the relationship between the soil properties and the result of the evaluation is poorly defined. The results from this type of system, perhaps more than any other, need careful validation by values for soils under practical use.

            Some systems pursue agricultural ends while others seek exclusively engineering uses (such as the support and placements for structures, roads, septic tanks, etc). Within the systems for agricultural uses, some, called land capability, evaluate the capacity of the soil for general use (crops, pasture, forestry), while others, called land suitability, evaluate the suitability of the soil for specific uses, such as a specific crop (e.g., wheat, potato) and with a particular kind of soil management. The classifications of land capability define the degrees of capacity in generally vague terms, focusing fundamentally on the limitations for the general use. Land suitability provides more practical results but need more data both for the land as well as for concrete specifications for each type of crop. An important compilation of data on the optimal and marginal conditions of many crops is provided for the tropical and subtropical regions by Sys et al. (1993).

Frequently the evaluation systems take into account the beneficial effects that can result from introducing certain improvements, these known as potential evaluations.

Evaluation systems use mapping methods to represent the results of the soil evaluation.

 

Major systems for agricultural soil capability evaluation

 

Many systems have been used to evaluate the quality of agricultural soils, the following systems being notable.

 

Land Capability Classification. This method was established by the Soil Conservation Service de USA according to the system proposed by Klingebiel and Montgomery (1961) and has been widely used throughout the world with numerous adaptations. It is a categorical system that uses qualitative criteria. The inclusion of a soil within a class is made in the inverse manner—that is, without directly analysing its capacity, but rather its degree of limitation with respect to a parameter according to a concrete use. Some factors that restrict soil use can be used to define the productive capacity (intrinsic: soil depth, texture, structure, permeability, rockiness, salinity, soil management; extrinsic: temperature and rainfall) and yield loss (slope of the terrain and degree of erosion). Five systems of permanent agricultural exploitation are considered: permanent soil cultivation, occasional soil cultivation, pasture, woods and natural reserves. This system seeks maximum production with minimum losses in potential. Three levels of classification ware established: classes, subclasses and units. Also, 8 classes with increasing limitations in use are defined from I to VIII. As a function of the permitted uses, 4 use groups can be distinguished: permanent soil cultivation (or any type of exploitation; Class I, suitable soils; Class II, good soils but with some limitations; Class III, soils acceptable but with severe limitations), occasional soil cultivation (pastures, woods or natural reserves; Class IV, not recommended for agricultural use for severe limitations and/or required careful management); no soil cultivation, only pastures (in forests or natural reserves; Classes V, VI and VII) and natural reserves (Class VIII).

            Depending on the type of limitation, various subclasses of capacity are established: e, for erosion risks; w, for wetness and drainage; s, for rooting and tillage limitations resulting from shallowness, drought risk, stoniness, or salinity; c, for climatic limitations. The capability units represent similar proposals of use and management.     

            Many authors have amplified the number of limiting characteristics used. Furthermore, it is frequent that in their application, quantitative criteria are introduced (Bibby and Mackney, 1969; Burnan and McRae, 1974; Bartelli, 1978) and have even changed the number of classes and subclasses defined (in Portugal, Azevedo and Cardosso, 1962; in Pakistan, Islam, 1966; in India, Murphy et al., 1968; in Ghana, Obeng, 1968; in England, Bibby and Mackney, 1969; in Nigeria, Carroll, 1974; in Spain, Sánchez et al., 1984).   

            This system presents unquestionable advantages, although it does not lack disadvantages. The classes are defined with criteria that are very general, simple and easy to understand as well as adapt to very diverse regions, but it proves difficult to apply with objective criteria. All the evaluation characteristics that make up the agricultural capacity have identical weight. The same class, with only one parameter (the maximum limiting factor) that classifies the soil within a certain class, embraces highly different soils. This system provides a highly general classification of soil capacity, since it dispenses with many soil characteristics of undeniable interest, but has the advantage of not requiring a detailed knowledge of the soil. On not using all the limiting factors that affect the use capacity of a soil, important information is lost for the user (Año et al., 1997). Its use proves quite subjective, though it adapts well to the experience of the evaluator. Its results materialize very well on a map, avoiding the erroneous evaluations that parametric methods can produce (McRae and Burnham, 1981).

 

Storie Index (1933). This represents the first parametric approach that was developed. It is an index that uses the multiplicative scheme. In addition, it uses intrinsic properties of the soils (genetic profile, parent material, profile depth, texture, drainage, nutrients, acidity an alkalinity), characteristics of the soil surface (slope and microrelief) and aspects of soil conservation (degree of erosion). The evaluation properties are grouped into four factors that are quantified in the corresponding tables. The factors are weighed a priori, the more important being related on a scale from 5 to 100 and the less important factors from 80 to 100.

            With this index, general agricultural soil uses can be evaluated (hence it is a soil-capability evaluation method). To formulate the index, the four factors are multiplied together and the index is expressed as a percentage. Six classes are defined at the degree level, with decreasing values from 1 to 6. The degrees 1 to 3 are for agricultural use, degree 4 for very limited agricultural use, 5 for pasture and 6 without use. Subdegrees are established according to limiting factors: “s” for depth, “p” for permeability, “x” for texture, “t “ slope, “d” for drainage and “a” for salts.

It is important to emphasize that this system does not consider climatic characteristics. Thus the evaluation is of the soil itself, valid for comparing the soils of a certain region with the same type of climate.

This evaluation index was developed for California, and thus application to other regions of the world has involved numerous modifications (in Canada by Bowser, 1940; in India by Shome and Raychaudhuri, 1960; in tropical countries by Sys and Frankart, 1972; in arid regions by Sys and Verheye, 1974).

 

Productivity index of Riquier et al. (FAO, 1970). The basic concept of this method is that agricultural-soil productivity, under optimal management conditions, depends on the intrinsic characteristics. This is a multiplicative parametric method to evaluate soil productivity, from a scheme similar to the Storie index. The concept of productivity is defined as the capacity to produce a certain quantity of harvest per hectare per year, expressed as a percentage of optimal productivity, which would provide a suitable soil in its first year of cultivation. The introduction of improvement practices leads to a potential productivity or potentiality. The quotient between the productivity and the potentiality is called the improvement coefficient.

            The evaluation is made for three general types of use: agricultural crops, cultivation of shallow-rooted plants (pastures), and deep-rooted plants (fruit trees and forestation).   

            The determining factors of soil depth are: wetness, drainage, effective depth, texture/structure, base saturation of the adsorbent complex, soluble-salt concentration, organic matter, cation-exchange capacity/nature of the clay and mineral reserves. The parameters of the soil surface (e.g., slope, erosion, flood tendency, or climate) are not considered      

            The different parameters are evaluated in tables and, as also occurs in the Storie index, the evaluation factors present different weights.

            Productivity is expressed as the product of all these factors expressed in percentages. Five productivity classes are defined: class P1 = excellent; class P2 = good, valid for all types of agricultural crops; class P3 = medium, for marginal agricultural use, suitable for non-fruiting trees; class P4 = poor, for pasture or forestation or recreation; class P5 = very poor or null, soils not adequate for any type of exploitation.

The improvement coefficient is the ratio between the productivity and the potentiality and represents a good index for evaluating the feasibility of these possible improvements

This is a quantitative method, precise (although the partial scoring of the different parameters is quite arbitrary), objective (the only objective is the mathematical method), simple and easy to calculate. The evaluations reflect the degree of suitability of the different evaluation parameters, so that it proves easy to determine the possible improvements for each soil. The evaluation parameters as well as the resulting assessment can be adapted to local conditions.

 

Soil Fertility Capability Classification (FCC). This was proposed by Buol et al., (1975) and modified by Sanchez et al. (1982) to evaluate soil fertility. In this system, three levels or categories were established. The first, the type, was determined by the texture of the arable layer, or of the first 20 cm, if this is thinner. Its denomination and range are: S, sandy (sandy and sandy loam); L, loams <35% clay (excluding sandy and sandy loam); C, clayey > 35% clay; O, organic > 30% organic matter to 50 cm or more.

The type of substrate is the second level and is used when there is a significant textural change in the first 50 cm of the soil. It is expressed with the same letters, adding “R” when a rock or a hard layer is found within this depth.

The third level is comprised of the modifiers, which are the chemical and physical parameters that negatively influence soil fertility. These are numerous and are represented by lower-case letters.

            In the denomination of the soil class, the principle limitations for use are directly represented. For example, for an Orthic Solonchak, the FCC class that represents it is LCds, which signifies that it is a soil susceptible to severe erosion (L), limited drainage (C), dry soil moisture regime (d) and with salinity (s).

 

Major systems for land agricultural suitability evaluation

 

            The methods of land suitability evaluate the aptness of soils for certain crops and specific management. These are doubtless the most logical systems, as each soil use has its own demands, while the evaluations of land capability, of general uses, are considered limitations with mean values that affect the most usual uses. In these specific evaluations, socio-economic factors take on great importance. It requires that the benefits exceed the necessary investment, and for the evaluation, the local or national determinants should be considered. The evaluation has two focuses: to determine either which lands are best for a certain crop or which crop is suitable for each type of land. These evaluation systems can be as numerous as the soil uses. To homogenize the criteria, the FAO has proposed an evaluation system which, more than a complete system, is a scheme with general guidelines to formulate specific evaluation systems.

 

The FAO Framework for Land Evaluation (1976). The FAO Framework for Land Evaluation (FAO 1976 and subsequent guidelines: for rainfed agriculture, 1983; forestry, 1984; irrigated agriculture, 1985; extensive grazing, 1991) is considered to be a standard reference system in land evaluation throughout the world (Dent and Young, 1981; van Diepen el al., 1991), and has been applied both in developed as well as developing countries.

            This framework is an approach, not a method. It is designed primarily to provide tools for the formulation of each concrete evaluation. The system is based on the following concepts: i) the land is qualified, not only the soil. ii) Land suitability must be defined for a specific soil use (crop and management). iii) Land evaluation was to take into account both the physical conditions as well as economic ones; iv) The concept of land evaluation is essentially economic, social and political. v) The evaluation requires a comparison between two or more alternative kinds of use. vi) The evaluation must propose a use that is sustainable. vii) A multidisciplinary approach is required (Purnell, 1979; van Diepen et al., 1991).

            In the scheme, four categories are recognized. The highest category is the order that it reflects, in broad features, whether a soil is suitable or not for a given use. Two orders are recognized:

S = Suitable. Land in which the benefits exceed the costs and sustained use does not incapacitate the soil over a sufficiently long period of time.

N = Not suitable. Land can be classified as not suitable for a certain use for diverse reasons. The use proposed may be deemed technically impractical, as in irrigation of abrupt rocky terrain, or that it causes serious environmental degradation, as in cultivation on steep slopes. Frequently, however, the reason is economic, in that the profit expected does not justify the cost required.              

            The second category is the class that reflects degrees of suitability within the order. These are numbered consecutively in Arabic numerals.

            For the order S, three classes are considered:

                        S1 = Highly suitable. Without limitations for sustained use or minor limitations that do not affect productivity nor appreciably increase costs.

                        S2 = Moderately suitable. Moderately serious limitations that reduce profits or involve risks of degradation in the sustained use of the soil.

                        S3 = Marginally suitable. The limitations for the sustained use are serious and the balance between the costs and benefits make the use only marginally justifiable. Its use is normally justified on other than economic grounds.

            In the order N, three classes are also recognized:

                        N1 = Not currently suitable. Land with limitations that could be eliminated by technical means or investment, but that these changes are at present unfeasible.

                        N2 = Permanently unsuitable. Serious limitations of generally a physical nature, which are assumed to be beyond solving over the long term.

                        X = Land for conservation. Unsuitable for exploitation, being lands of special protection, due to their conservation, wildlife, of special scientific, ecological or social interest (e.g., parks, reserves or recreational zones).

            The limits between the orders (S and N) and between the different classes (S1, S2, S3 and N1, N2) are established by the presence of limiting factors. One limiting factor is a characteristic of the soil that hampers its use, reduces productivity, increases costs and implies degradation risk, or all of the above.

            These limiting factors are used to define the third category of the system, which is the subclass. In the symbol of each subclass, the number of limitations involved should be kept to the minimum one letter, or, rarely, two. The limitations proposed include: t, slope; e, erosion risk; p, depth; s, salinity; d, drainage; c, bioclimatic deficiency; r, rockiness; i, flood risk.

            Finally, the fourth category is the unit that establishes the differences within the subclasses as a function of the desired use. All of the units within a subclass (S2rA, S2rM, ...) have the same degree of suitability at the subclass level (S2) and analogous characteristics of limitation at the subclass level (r). The units differ from each other in their characteristics of production or in secondary aspects of their management demands. Their examination enables a detailed interpretation at the planning level of the exploitation. The units are distinguished by upper-case letters that are placed at the end. There is no limit at all for the number of units examined within a subclass. These defined are: A, intensification in the agricultural use without need of great improvements; M intensification in the agricultural use with need of major improvements (irrigation, etc.); P, use for pasture for livestock; F, forestation.

In some cases, the designation “conditionally suitable” can be added so long as certain conditions are satisfied.

A very important role in this framework is represented by the concept of Land Utilization Type (LUT). It represents a thorough description of soil use for a proposed use, in terms of crop (type and rotation), soil management (working of the soil, additives and possible irrigation) and socio-economic framework (labour cost, market, distribution, expenses, profits and subsidies). To propose a certain LUT, the crop must meet certain requirements of the land that satisfy the LUT and these are grouped under the concept of Land Use Requirements (LUR). For these requirements to be fulfilled, the land must have certain features, known as Land Qualities (LQ), which are supported by particular characteristics of the land, called Land Characteristics (LC). The LCs represent characteristics of the soils that can be examined and measured in the profile or in the laboratory, while the LQs are complex qualities that are not directly measurable but rather are estimated from a certain combination of LCs. The LQs are a direct result of the LCs. The LQs used for defining the LUT are numerous (although much less than the LCs) and as diverse as nutrient availability, workability, flooding hazards, resistance to degradation, or accessibility of the terrain.

In addition, each of these LQs are based on a set of LCs, and thus, for example the LQ “nutrient availability” is a combination of the following LCs: organic matter, nutrients (N, K, P), cation-exchange capacity, degree of saturation, pH, salinity, carbonates, depth and gravel. Meanwhile, the LQ “water availability” is based on the following LCs: texture, structure, drainage, available water and water reserve. In our opinion, to facilitate the elaboration of a certain LUT, it might be useful to use the support of LQ organized as soil qualities (SQ), climate qualities (CQ), topography qualities (TQ), management qualities (MQ) and socio-economic qualities (EQ), which are calculated on the basis of their corresponding characteristics SC, CC, TC, MC and EC.

Altogether, the Framework FAO represents a highly useful and flexible system which is easily adapted to local characteristics. The main drawbacks are: i) the conceptual confusion caused by using pre-existing terms; ii) the poor distinction between the physical, social and economic characteristics.

An evaluation requires: i) the selection and description of the LUTs; ii) definition of the necessary LQs on the basis of the LCs; iii) definition of the suitable values of the LCs; iv) measurements of the LCs in the land units; v) an economic and social study; vi) presentation of the final suitability evaluation and reflection on a map with the corresponding evaluation units. For the LUTs, LQs and LCs, 5 conditions are established according to suitability (S1, S2, S3, N1 and N2). As the differentiating feature, the greatest limitation alone may be used. A LUT would be defined as pertaining to class S2ab if it presented the LQa and LQb of level S2 and the LQc of level S1. Or a combination of limitations (number and intensity) can be taken. The previous LUT would be defined as belonging to class S3 for having two LQs with an S2 value. Also, a correction factor can be introduced according to the soil depth at which the limitations are found.

The characteristics to take into account when making a land evaluation can be chosen by means of a statistical treatment. For example, to identify which land characteristics have repercussions in the production of a certain crop, a principal-components analysis can be performed, followed by an analysis of variance (ANOVA) to differentiate the significant characteristics. The simple regression highlights the significant properties most directly related to production. Finally, the establishment of the multiple-regression equation can establish the values of the characteristics chosen that define the suitability levels (S1, S2, S3 and N).

Boixadera and Porta (1991) calculated land capability by beginning with a series of values for land suitability for different crops. They developed a “value index” for the Ministry of Economy and Internal Revenue of Spain, making an evaluation for 8 LUTs (corn, wheat, tomato, lettuce, potato, peach, apple and alfalfa) with two management intensities. The value index represents the mean of 16 individual evaluations.

Aguilar and Ortiz (1992) used the FAO Framework, combining it with the parametric Riquier index to define the suitability classes (S1, S2, S3, N1 and N2). Furthermore, the classes were defined for land capability instead of for land suitability as proposed in the FAO Framework.

A land evaluation can be made with the FAO Framework automatically with the useful computer program ALES (Rossiter and van Wambeke, 1995), which enable an easy construction of an expert system for land evaluation. This is not an expert system in itself, as it provides no information, but rather is an empty system that enables evaluators to introduce their own evaluation characteristics (choose their own LUTs, LQs and LCs), according to their own knowledge and data availability. The MicroLEIS of De la Rosa et al. (1992) is a complete computer program to evaluate according to the FAO Framework adapted to the Mediterranean environment.

 

The Irrigation Suitability Classification of the Unites States Bureau of Reclamation (USBR, 1953). The estimation of the land capacity for irrigation is basic in the plans for development, as irrigated crops constitute the most productive form of agriculture and are especially profitable in arid and semi-arid regions.

            The characteristics and qualities of the soil, needed in the evaluation related to irrigation, concern soil, drainage, hydrology, topography, vegetation, technical problems, economics, and social as well as political matters. Therefore, regional planning of an irrigation project requires multidisciplinary solutions.

            This is a classic system, widely used, which is based on the correlations between the different factors affecting production under irrigation. The consideration of economic determinants, as a starting basis, gives rise to more realistic capacity classes. The variability of the transformation is measured by the capacity for potential payment with a view towards the amortization of the project, the maintenance of long-term productivity of the land, preventing erosion, the degradation of the structure, salinity and the continuous flooding. The payment capacity compares the costs of transformation and production against the potential productive capacity. This latter concern is a function of climate, soil, topography (slope, relief, position), quantity and quality of the irrigation water, and the drainage system.

            The Land Class is defined as a “category of land having similar physical and economic attributes which affect the suitability of land for irrigation” (McRae and Burnham, 1981). The USBR system establishes six classes to evaluate the suitability of the soils for irrigation. The parameter used and its ranges are reproduced in the corresponding tables. To facilitate the reading of the evaluation maps on each cartographic unit, a formula is written in which all the representative data are reflected.

 

Comparison of the four systems of soil-capability evaluation

 

The results have been compared (using MicroLEIS software) on applying four systems of soil-capability evaluation to the same soil: Land Capability Classification (LCC), Storie Index (SI), Riquier Productivity Index (RPI), and the FAO Framework (FK). With all these systems, quantitative parameters were used. The equivalencies estimated for the classes established by these evaluation systems are presented in Table 2.

 

 

Table 2. Comparisons between the classes defined by the soil-evaluation systems.

LCC, Land Capability Classification; Si, Storie Index; RPI, Riquier Productivity Index; FK, FAO Framework.

 

A selection of 30 soils in Andalusia (S Spain) with different characteristics (De la Rosa, 1984; Sierra et al., 1986; Aguilar et al., 1986) have been evaluated. The results of the evaluation are presented in Table 3.

The results show matches between the evaluations by, on the one hand, the systems LCC and FK, with respect to SI and RPI on the other. The two former systems appear to offer, for the soils evaluated, more real results than those offered by the latter two.

In broad terms, the results show a certain divergence. Of the 30 soils evaluated, only one complete correspondence appears in 9 cases (soils 2, 3, 5, 8, 18, 21, 22, 23 and 26). Thirteen soils showed a deviation of no more than one jump of class for each evaluation system (soils 6, 7 10, 11, 12, 13, 15, 17, 24, 27, 28, 29 and 30). The best equivalencies were presented when the maximum limiting factor was soil depth (soils 8, 18, 22, 23 and to a minor degree 6, 11, 12 and 17). With the exception of the evaluations by the RPI method, the limiting factor slope also gave coherent evaluations (soils 1, 21 and to a lesser extent also 14, 15, 20 and 25).

Major divergences appeared between the evaluations in 8 soils, the most striking cases being: i) The RPI method, which on occasions gave anomalous results for not using the parameter slope in the calculations (soils 19, 20 and 25) and on other occasions the anomaly did not appear after offsetting the absence of this parameter with the presence of other limiting factors (soils 5, 15 and 21); ii) The SI system, which presented some apparently erroneous results for not taking into account the parameter moisture (soil 16); iii) The SI, in which the presence of abundant stones, pebbles and gravels was apparently hardly penalized (soils 4 and 19); iv) the RPI, which gave a lower degree of class than did the other three systems in the case of soil 1, but we incline to regard this as the correct assessment, as this soil had a low temperature regime, and the other three methods did not evaluate this possibility ( considerations analogous to the situation with soil 14).

 

Soil type	Parent material	LCC	SI	RPI	FK
1 Typic Cryosaprist	micaschist	IVsp	4ps	P5fp-->(P3)	S3sp
2 Typic Xerofluvent	alluvial	II	2	P2	S2
3 Typic Xerofluvent	alluvial	I	1	P1	S1
4 Typic Xeropsamment	dolomite	VIIr	3g-->(6)	P5g	S3r-->(N)
5 Lithic Xerorthent	micaschist	VIIs	6dg	P5dg	Ns
6 Lithic Xerorthent	dolomite	VIgr	5dgr	P5dg-->(P4)	S3d-->(N)
7 Typic Chromoxeret	marl	II	3p-->(2)	P2p	S3-->(S2)
8 Calcixerollic Xerochrept	marl	IVd	4d	P3d	S3d
9 Calcixerollic Xerochrept	sandstone	VIg	4gd-->(5)	P5g-->(P4)	S2-->(N)
10 Calcixerollic Xerochrept	conglomerate	III	3d	P2-->(P3)	S2
11 Lithic Xerochrept	slate	IIId-->(4)	5dr-->(4)	P3d	S3d
12 Lithic Xerochrept	granite	IId	3d-->(2)	P2dt	S2
13 Typic Humaquept	micaschist	Vp	5p	P5p-->(P4)	S3pf-->(N)
14 Typic Cryumbrept	micaschist	IIIs-->(IV)	4s	5fg-->(P3)	S3sf
15 Typic Haplumbert	micaschist	VIIs	5sg-->(6)	P5gf	Ns
16 Vertic Haplargid	andesite	Vm	2-->(5)	P5m-->(P4)	Nm
17 Petrogypsic Gypsiorthid	silts, gypsum	IVdg	5dg-->(4)	P4dg-->(P3)	S3d
18 Lhitic Haploxeroll	conglomerate	VIIrd	6d	P5dg	Nd
19 Calcic Haploxeroll	micaschist	VIIs	4dg-->(6)	P2-->(P5)	Ns
20 Typic Haploxeroll	sandstone	VIIs	5sg-->(6)	P2-->(P5)	Ns
21 Typic Haploxeroll	micaschists	VIIs	6s	P5gf	Ns
22 Udic Haplustoll	serpentine	IIId	3dt	P3d	S2
23 Mollic Haploxeralf	limestone	IVd	4d	P3d	S3d
24 Typic Haploxeralf	slate	IIIe	3e	P2-->(P3)	S3e
25 Xerochreptic Haploxeralf	slate	IIIs	3se	P1-->(P3)	S3se
26 Typic Rhodoxeralf	conglomerate	I	1	P1	S1
27 Calcic Rhodoxeralf	conglomerate	IIg	1-->(2)	P1-->(P2)	S2m
28 Mollic Palexeralf	limestone	IIIr	3t	P2t-->(P3)	S2
29 Typic Palexerult	slate	IIIs	3r	P2-->(P3)	S2
30 Typic Palexerult	clays	IIes-->(III)	3t	P3t	S2

Table 3. Evaluation of 30 soils by four methods of soil-capability evaluation.

LCC, Land Capability Classification; SI, Storie Index; RPI, Riquier Productivity Index; FK, FAO Framework.

Limiting characteristics: e, erosion; d, depth; g, gravels; f, frozen; m, moisture; p, permeability or drainage or flooding; r, rocks or pebbles or stones; s, slope; t, texture or structure.

In bold, the results that do not coincide with the evaluations of the other methods; in parenthesis the results that would correspond with the other methods. In bold and cursive, results that strongly differ from those of the other methods.

 

Trends and perspectives

           

The foregoing results indicate heterogeneity in the soil assessment by the different soil-evaluation systems. It would be necessary for the evaluation made for a soil and specific end (for example, agriculture) to give consistently the same capability class, regardless of the system used. Thus, it appears to be advisable for the evaluator of the soils, before developing a project of soil evaluation, to make a comparative study of three or more evaluation systems in order to choose the system that best matches the objectives set, and, above all, to calibrate the system, modifying the characteristics used for the evaluation and readjusting the values.

Here, we note the generalized opinion that in land evaluation an assessment should be made not only of the characteristics of the physical medium but also of the social, economic and political ones (Aandhal, 1958, Bie et al., 1973; Sys, 1985; van Diepen el al., 1991, Rossiter, 1995 and 1996). In addition, there appears to be agreement among authors in that land evaluation should use quantitative data and results—the more quantitative, the better (Bouma, 1989; Burrough, 1989; van Diepen el al., 1991; Hoosbeek and Bryant, 1992; Rossiter, 1996). The qualitative systems could be valid to make soil evaluations for the first time in a region and afterwards develop the quantitative land evaluation in the most suitable areas.

In addition, it seems correct to state that the projects of soil and land evaluation are steadily more oriented towards suitability than towards capability.

The results of soil and land evaluation should be expressed in a simple manner to be easily understood by non-specialists. Also, we deem it important for the evaluations of each land unit to offer several prescriptions of use in order to facilitate acceptance by farmers and planners.

Rossiter (1996), in his perceptive paper concerning the framework for land evaluation, established three modalities: 1) Non-spatial models of single-area land suitability; 2) spatial models of single-area land suitability; 3) models of multi-area suitability and the land-allocation problem. This systematic analysis represents a comprehensive overview of all existing methods of land-evaluation frameworks and it places each method in the same framework, expressed according to mathematical formulas.

Some authors underline the importance of dynamic evaluations as opposed to static ones in current methods (Wosten and Bouma, 1985; Bouma and van Lanen, 1986; Bouma, 1989; McBratney and Whelan, 1995; Rossiter, 1996). Thus, van Diepen et al. (1991) point out the possibility for land characteristics to be calculated in specific periods, such as before, during, and after the growing season.

Sustainability is a concept that is becoming more and more accepted among planners. In this sense, the FAO (Smyth and Dumarski, 1993) has developed a worthwhile International Framework for Evaluation Sustainable Land Management (FESLM), in which sustainability is defined as “a measure of whether or not a defined system of land use can be maintained at acceptable levels of productivity or service with realistic levels of input yet without progressive physical, biological, economic, or social damage to the environment on a specific site over a stated period of time”. The classification relies on “indicators” (highly unstable characteristics that change predictably with respect to a certain environmental pressure), “thresholds” (limit at which the change in the indicator becomes apparent), and “criteria” (standards or rules, such as models, tests or measurements, which govern judgements on environmental conditions).

            To date, in land-evaluation projects, production has been valued over all other concerns. We feel that the time has arrived to concede the importance of environmental and human-health issues, as well as sustainability (FAO, 2001).

Rossiter (1996) highlighted the need to adapt land evaluation to the demand of the clients and not, as previously, with an exclusively scientific perspective.

Also, it would be useful provide the degree of uncertainty in these studies, and thereby reinforce the credibility of the use recommendations (Rossiter, 1996).

Land evaluation is a tool for strategic land use planning. There are many models for simulation and many computer packages for application of land evaluation in land-use planning (McRae y Burnham, 1981; FAO, 1993). The most widely used models are the descriptive ones, called the “black box”, in which the inputs are related to the results, frequently by mathematical procedures, without attempting to explain the scientific bases involved. The prime problem is in applying the model to other regions aside from the one for which the system was calibrated, frequently with local data outside the range of values for which the program was developed. Gradation matrices, decision trees and neuronal networks, together with mathematical equations are the supports on which these models are most often constructed. The statistical models are static and at times do not offer good results, requiring the use of dynamic models (time-dependent) which consider annual and interannual variations.

Explicative models seek to explain the way in which systems function. These are the “white box”, in which an attempt is made to explain the results in view of the input data on the basis of scientific knowledge of the processes that have taken place. These are still in their preliminary phases, although they have a promising future. Today, they lack adequate scientific support, as the natural world is too complex to be synthesized in a necessarily limited number of variables, and therefore this system needs a thorough calibration and validation with extensive casuistry.

Remote sensing techniques provide important spatial and temporal information, and, when their features are improved, they may be a valuable tool for land evaluation.

The Framework FAO for land evaluation, despite the time which has passed since its appearance, continues to be the most widely used system, and it represents the standard land-evaluation system, although its terminology is at times vague and causes frequent confusion (especially the concept of land quality). Other times, the errors are due to incorrect applications of the Framework. An advance in the knowledge of the interrelationships between the LCs would improve the definition of the LQs, while a better relationship between the physical, technical, social and economic characteristics will enable easier use and appreciably improved recommendations in land evaluation.

 

References

 

Aandhal, A.R. 1958. Soil survey interpretation-theory and purpose. Soil Science Soc. Amer. Proc. 22:152-154.

Aguilar, J., Simon, M., Fernandez, J., Gil de Carrasco, C. and Marañes, M. 1986. Aldeire-1028. Memoria del Mapa de Suelos. Proyecto LUCDEME. Ministerio Agricultura. ICONA. Universidad de Granada. Spain.

Aguilar, J. and Ortiz, R. 1992. Metodología de capacidad de uso agrícola de los suelos. III Congreso Nacional de la Ciencia del Suelo. 281-286. Pamplona. Spain

Año, C., Sánchez, J., and Antolín, C. 1997. Análisis y valoración de los sistemas de evaluación de suelos en España. Evolución, tendencias actuales y perspectivas futuras. Estudios Geográficos. LVII, 228, 331-353.

Azevedo, A.L. and Cardoso, J.C. 1962. Soil classification in Portugal and its application in agricultural research. Trans. Comm. IV and V Int. Soc. Soil Sci. New Zealand, 473-479.

Bartelli, L.J. 1978. Technical classification systems for soil survey interpretation. Adv. Agron. 30:247-289.

Bibby, J.S. and Mackney, D. 1969. Land Use Capability Classification. Soil Surv. Gt. Br. Tech. Mono. No. 1.

Bie, S.W., Ulph, A. and Beckett, P.H.T., 1973. Calculating the economic benefits of soil survey. Journal of Soil Science, 24: 429–435.

Boixadera,J. and Porta, J. 1991. Información de suelos y evaluación catastral. Método del Valor Índice. Ministerio de Economía y Hacienda. Madrid.

Bouma, J. 1989. Using soil survey data for quantitative land evaluation. In: Advances in Soil Science (ed. B.A. Stewart). Springer, New York, pp. 177-213.

Bouma, J. and van Lanen, H.A.J. 1986. Transfer functions and threshold values: from soil characteristics to land qualities. In Quantified land evaluation procedures: Proceedings of the international workshop on quantified land evaluation procedures held in Washington, DC 27 April - 2 May 1986. Washington, DC: ITC.

Bouma, J. and Bregt, A.K. (ed.) 1989. Land qualities in space and time. Proceedings of a symposium organized by the International Society of Soil Science (ISSS), Wageningen, the Netherlands 22-26 August 1988. Pudoc, Wageningen.

Bowser, W.E. 1940. Soil surveys in relation to land classification in Alberta. Scient. Agric. 20: 285-290.

Buol, S.W., Sanchez, P.A., Cate, R.B. and Granger, M.A. 1975. Soil fertility capability classification: a technical soil classification system for fertility management. In Bornemisza, E. and Alvarado A. (Ed.) Soil Management in Tropical America. N.C. State Univ., Raleigh, NC: 126-145.

Burnham, C.P. and McRae, S.G. 1974. Land judging. Area, 6: 107-111.

Burrough, P.A., 1989. Matching spatial data bases and quantitative models in land resource assessment. Soil Use and Management, 5: 3–8.

Carroll, D.M. 1974. The soils of the Maiduguri-Bama area. Soil Surv. Bull. Samaru 40.

De la Rosa, D. 1984. Catalogo de suelos de Andalucía. Junta de Andalucía. Agencia de Medioambiente. Sevilla. Spain.

De la Rosa, D., Moreno, J.A., Garcia, L.V. & Almorza, J. 1992. MicroLEIS: A microcomputer-based Mediterranean land evaluation information system. Soil Use & Management 8, 89-96.

Dent, D. & Young, A. 1981. Soil survey and land evaluation. George Allen & Unwin, London.

FAO 1976. A framework for land evaluation. Soils Bulletin 32. Food and Agriculture Organization of the United Nations, Rome, Italy.

FAO 1983. Guidelines: land evaluation for rainfed agriculture. Soils Bulletin 52. Food and Agriculture Organization of the United Nations, Rome, Italy.

FAO 1984. Land evaluation for forestry. Forestry paper 48. Food and Agriculture Organization of the United Nations, Rome, Italy.

FAO 1985. Guidelines: land evaluation for irrigated agriculture. Soils Bulletin 55. Food and Agriculture Organization of the United Nations, Rome, Italy.

FAO 1991. Guidelines: land evaluation for extensive grazing. Soils Bulletin 58. Food and Agriculture Organization of the United Nations, Rome, Italy.

FAO. 1993. Computerized systems of land resources appraisal for agricultural development, World Soil Resources Report No 72. Rome. Italy.

FAO 2001. Indicadores de la calidad de la tierra y su uso para la agricultura sostenible y el desarrollo rural. Boletín de tierras y aguas de la FAO No 5. Food and Agriculture Organization of the United Nations, Rome. Italy.

Hoosbeek, M.R. & Bryant, R.B. 1992. Towards the quantitative modeling of pedogenesis - a review. Geoderma 55, 183-210.

Islam, A. 1966. Current status of soil and land capability classification in East Pakistan. Cento. Conf. Ld. Classif. non-irrig. Lds. 81-84.

Klingebiel, A.A. & Montgomery, P.H. 1961. Land capability classification. USDA Agricultural Handbook 210. US Government Printing Office, Washington, DC.

McBratney, A.B. and Whelan, B.M., 1995. Continuous models of soil variation for continuous soil management. pp. 325–338. In P.C. Robert, R.H. Rus., and W.E. Larson (Editors), Site-specific Management for Agricultural Systems. American Society of Agronomy/ Crop Science Society of America / Soil Science Society of America, Madison, Wisconsin.

McRae, S.G. & Burnham, C.P. 1981. Land evaluation. Monogr. soil survey. Clarendon Press,

Murphy, R.S., Jain, S.P. and Naga B.S.R. 1968. Soil survey and land capability classification for sound watershed management in Kundah Project (Madras). J. Indian Soc. Soil Sci. 16: 223-227.

Obeng, H.B. 1968. Land capability classification of the soils of Ghana under practices of mechandized and hand cultivation for crop and livestock production. Trans. 9th Int. Cong. Soil Sci. 4: 215-223.

Purnell, M.F. 1979. The FAO approach to land evaluation and its application to land classification for irrigation. Wld. Soil Resour. Rep. 50:4-8.

Riha, S.J., Rossiter, D.G. & Simoens, P. 1994. GAPS: General-purpose Atmosphere-Plant-Soil Simulator. Version 3.0 User's Manual. July 1994 ed. SCAS Teaching Series. Cornell University, Department of Soil, Crop & Atmospheric Sciences, Ithaca.

Riquier, J., Bramao, L. and Cornet, S.P. 1970. A new system or soil appraisal in terms of actual and potential productivity: FAO Soil Resources No 38. Rome. Italy.

Rossiter, D.G. & Van Wambeke, A.R. 1995. Automated Land Evaluation System: ALES Version 4.5 User's Manual. December 1994 printing. SCAS Teaching Series No. T93-2, Revision 5. Cornell University, Dp of Soil, Crop & Atmospheric Sciences, Ithaca, NY.

Sánchez, P.A., Couto, W. & Buol, S.W. 1982. The fertility capability soil classification system: interpretation, applicability and modification. Geoderma 27(4): 283-309.

Sanchez, J., Rubio, J.L., Martinez, V. and Antolin, C. 1984. Metodología de capacidad de uso de los suelos para la cuenca mediterránea. I C. N. Ciencia Suelo. 937-948. Madrid. Spain.

Shome, K.B. and Raychaudhuri, S.P. 1960. Rating of soils of India. Proc. Nant. Inst. Sci. India 26a suppl. I:269-289.

Sierra, C., Ortega, E., Garcia, I., Rodriguez, T. Saura, I. and Iriarte, A. 1986. Durcal-1041. Memoria del Mapa de Suelos. Proyecto LUCDEME. Ministerio Agricultura. ICONA. Universidad de Granada. Spain.

Smyth, A.J. and Dumanski, J. 1993. FESLM: An international framework for evaluating sustainable land management. World Soil Resources Report No 73. Food and Agriculture Organization of the United Nations. Rome. Italy.

Storie, R.E. 1933. An index for rating the agricultural value of soils. Bulletin - California Agricultural Experiment Station. Vol. 556, University of California Agricultural Experiment Station, Berkley, CA.

Sys, C. and Frankart, R. 1972. Land capability in the humid tropics. Sols afr. 16: 153-175.

Sys, C. and Verheye, W. 1974. Land evaluation for irrigation of arid regions by use of the parameter method. Trans. 10th Int. Cong. Soil Sci. X: 149-155.

Sys, C. 1985. Land evaluation. State University of Ghent, International Training Centre for post-graduate soil scientists; Algemeen Bestuur van de Ontwikkelingss. Ghent, Belgium.

Sys, C., Van Ranst, E., Debaveye, J. & Beernaert, F. 1993. Land evaluation, Part 3 : Crop requirements. Agricultural Publications 7. General Admin. Develop. Coop., Brussels.

Rossiter, D. G., 1995. Economic land evaluation: why and how. Soil Use Mang. 11: 132–140.

Rossiter, D. G. 1996. A theorethical framework for land evaluation. Geoderma 72. 165-190.

USBR Department of the Interior Bureau of Reclamation 1953. Irrigated land use, Part 2: Land classification. B. R. Manual. Vol. 5, U.S. Government Printing Office, Washington.

van Diepen, C.A., Van Keulen, H., Wolf, J. & Berkhout, J.A.A. 1991. Land evaluation: from intuition to quantification. In: Advances In Soil Science (ed. B.A. Stewart). Springer, New York, pp. 139-204.

van Diepen, C.A., Wolf, J., van Keulen, H. & Rappoldt, C. 1989. WOFOST: a simulation model of crop production. Soil Use & Management 5, 16-24.

van Latesteijn, H.C. 1998. A policy perspective on land use changes and risk assessment. Agriculture, Ecosystems and Environment 67: 289–297

Wosten, J.H.M. and Bouma, J. 1985. Using simulation to define moisture availability and trafficability for heavy clay soil in the Netherlands. Geoderma 35: 187-196.