Chapter 2 - The socioeconomic attributes of trees

What are socioeconomic attributes?

Our aim in this chapter is to arrive at an understanding of the socioeconomic attributes of trees as they relate to the practical question of species choice in project planning. Before we can approach that question, however, we may need to spend a little time making sure we know how to recognize a socioeconomic attribute when we see one.

Intrinsic vs. ascribed attributes

We can begin by distinguishing straightaway between "intrinsic" and "ascribed" attributes. That trees have inherent biophysical attributes hardly anyone would deny. Some socioeconomic attributes, however, are wholly ascribed to trees by people, while others have a hybrid character reflecting the prevailing socioeconomic interpretation of intrinsic biophysical attributes. Still others are a manifestation of latent biophysical attributes that only reveal themselves under human management. Let us explore these differences through some examples.

All through the ages and in every land, trees have been considered fit receptacles for all manner of symbolism: the "Tree of Life," the "Tree of Knowledge," the "Speaking Tree," the "Hanging Tree," etc. Religious symbolism is one source of examples of what we mean by ascribed socioeconomic attributes, and they are important to the question of species choice to the extent that people's perception of specific trees in terms of these ascribed attributes casts an influence over their use (or non-use).

Thus, the same tree may have very different meanings to different people. For example, the "sausage tree" (Kigelia africana) is valued by the Wakamba of eastern Kenya as the source of an important ingredient for beer brewing, while among the Luo tribe in western Kenya--where the sausage-shaped fruit is used in funeral rites as a symbolic substitute for the body of a relative who has met a violent death in a distant place and whose remains could not be recovered--the tree is devoid of positive associations (Raintree and Hoskins 1988).

In Senegal some plants are classified as "brothers" because they foster each other's growth, while others are classified as "co-wives" because they inhibit each other's growth. Among the Bhil ethnic group in India, where polygyny has a different social meaning, "co-wives" are plants that grow well together (Hoskins 1987). These are examples of the use of analogies with human social relations to characterize the phytosociological behaviour of plants. The observations of plant behaviour may be quite accurate, and this is one way of encoding such knowledge in a culturally transmissible form, but as these two examples show, the appropriateness of a particular analogy may be limited to a particular ethnic group.

These examples illustrate the "hybrid" nature of many of the attributes ascribed to trees, where objective biophysical attributes are given a socially subjective interpretation. For our purposes, however, the most important socioeconomic attributes of trees are the ones that attach to trees according to their uses and accessibility to different users.

One of the most powerful sets of influences on tree use are those arising from the legally ascribed attributes of trees. Wherever there exists the possibility of legal sanctions involving fines or even imprisonment for violation of the forest code by cutting an "illegal" tree, all other attributes of the tree may pale into insignificance next to this legal attribute. The proscription against harvesting all trees of certain species, even when the potential user has planted them from his own nursery stock on his own land, is a notorious constraint on social forestry projects in francophone West Africa (Thomson 1987).

Legal attributes may encourage planting certain trees. In Gujurat, where the normal proscription forbidding harvest of trees was suspended for three trees--eucalyptus, leucaena and casuarina--it is said to be risky for people to plant any other species, regardless of whether or not these three trees are capable of meeting all of their needs for a variety of tree products (Saxena 1988). Governments seek to exert control over people's use of trees but their efforts can backfire, as in the case of eucalyptus planting in parts of India where this tree became a symbol and a rallying point of grass roots resistance to government meddling.

Do trees really have socioeconomic attributes? It is obvious that trees have intrinsic biophysical attributes and that these attributes make them more appropriate or less appropriate to different uses. It is also obvious that the different uses of trees have different degrees of relevance to different users and that the socioeconomic attributes of the individual user (as conditioned by his or her position within a particular social structure) must somehow influence and set limits on the relevance of particular trees.

Let us posit a set of relationships between the biophysical attributes of trees which make them suitable for certain uses, on the one hand, and the socioeconomic attributes of tree users on the other. It is not very clear at this point what precisely these relationships might be in any particular case, or what intervening variables might come into play to condition them, but it is clear that this whole complex refers to what we might usefully think of as the "socioeconomic attributes" of trees. For the purposes of this study we are most concerned with those attributes of trees that make them useful or useless, adoptable or non-adoptable, beneficial or harmful, relevant or irrelevant, etc., to different users in different socioeconomic settings.

The relativity of attributes

It follows that the same tree--indeed the same biophysical attributes of a single tree--may have different socioeconomic attributes or meanings for different people. What better way to illustrate the social relativity of tree attributes than to review the various ways in which the choice of a particular tree species could be "wrong" in the context of a particular community? We can see immediately that a species choice could be: 1) wrong for the community as a whole, or 2) wrong only for certain segments of the community.

The choice of a tree species is most likely to be wrong for an entire community when a project starts out with firm preconceptions about the kind of tree that is needed without first consulting with the community on its own perceptions of need. For example, riding on the bandwagon of funding for "energy crisis" issues, many community forestry projects were planned with fuelwood as the primary objective. But the project evaluation literature shows time and again that fuelwood is rarely rural peoples' first priority and that people are often more motivated to plant trees by the prospect of fruit, fodder, medicine or poles for cash.

The choice of tree is wrong only for certain segments of the community when the adoption of the tree planting practice is impossible, irrelevant or harmful not to the community as a whole but only to certain social categories of people within the community, e.g., the poor, the landless, herders, labourers, common property resource users, women, minority ethnic groups, etc.

Can we be more specific about this? In what specific ways could a species choice be wrong for certain categories of tree users? It is clear that any tree that is recommended for planting can be inappropriate for a particular client group in at least five different ways:

1) Inappropriate function: the tree isn't relevant to the client group's felt needs or general production objectives; e.g., firewood trees where there is no perceived fuelwood shortage, soil fertility improving trees where soil fertility is not felt to be a problem, timber trees instead of fodder trees for pastoralists, fodder trees where the natural supply is judged to be sufficient, etc.

2) Inappropriate market orientation: e.g., a cash crop tree where trees are needed primarily for subsistence purposes; a cash crop tree where transport costs or lack of marketing infrastructure make it impossible to turn a profit; or, at the other extreme, yet another subsistence tree when what is needed to stimulate more tree planting is a cash crop tree, etc.

3) Inappropriate land requirement: the tree is inappropriate to the spatial possibilities of the local landscape; e.g., attempting to introduce a tree best suited for monocrop block plantations in a situation where land is too scarce to be set aside for single purpose plantations and where the predominant land use is strongly oriented toward various forms of intercropping; or, vice versa, a tree specifically chosen for its compatibility with other crops in a situation of abundant land where there is no reason for not choosing a more productive "competitor" ideotype.

4) Inappropriate labour or skill requirement: e.g., a tree that requires intensive management in a situation where the labour or skills for this are unavailable or are otherwise occupied with more pressing demands; or a tree that requires too little labour in a densely populated area where labour absorption/employment creation is desirable.

5) Inappropriate capital requirement: trees that require special capital inputs that are unlikely to be available; e.g.., leguminous tree species that only perform well when inoculated with special strains of rhizobium that are unlikely to be made available locally; or special harvesting or processing equipment that is unavailable or too expensive.

These considerations give us a general idea of the range of concerns entailed by our study. Let us now sharpen our focus by reviewing briefly how the attributes of trees have been treated in the plant science literature.

Attributes, ideotypes and specifications

First developed by Donald (1968) and later elaborated by Donald and Humbling (1976) and others for use in the breeding of agronomic crops, ideotype literally means "a form denoting an idea" (Dickmann 1985). Thus far, the ideotype concept has been used primarily by breeders to define a plant model, which then becomes the target of a breeding programme. An ideotype specifies the ideal attributes of a plant for a particular purpose.

There is nothing in the powerful ideotype concept to limit the range of purposes that are considered, but in practice the application of the concept often evinces disciplinary biases that limit the range of purposes that are in fact considered. For example, forest ecologists and silviculturalists have been primarily interested in the way individual trees behave within forest stands and have used the ideotype concept mainly to advance the selection and breeding of trees for industrial forest plantations. This focus is made explicit in the explanation of the ideotype concept given by Dickmann:

Breeders select and breed individual trees and, at the start of any tree improvement programme, they must ask this question: in a particular environment (or on a certain type of site), using a prescribed silvicultural system, and assuming a well-defined end use of the harvested trees, what are the precise traits of a tree of the species under consideration that will produce a yield that approaches maximum? The answer to this rather complex question is a model tree, an 'ideotype,' and is the first step towards bioengineering an improved tree species.... In its broadest sense, an ideotype is a plant model that will yield a greater quantity of useful products than conventional cultivars or wild plants. The formulation of an ideotype is a truly practical step, because it provides a clear, workable goal toward which plant breeders can work (Dickmann 1985).

An example of this type of application is the ideotype described in Table 1, which was developed by foresters for industrial plantations in Finland.

Table 1. The conifer 'crop' tree ideotype for Finland (Larki and Tiberstedt 1985).

_____________________________________________________________

Stems

fast growing (in height & diameter) straight

small taper

thin bark

good quality timber

Branches

slow growing

short

small diameter

subtend angles of about 900 with the stem few flowers

large leaf areas per unit branchwood weight persistent foliage

produce long, green, narrow (columnar or pendulous) crowns

Other

wind firm

tolerant of snow loads

_____________________________________________________________

The management system to which this ideotype relates is not specified in much detail because it can be assumed that anyone interested in the ideotype will be familiar with standard silvicultural practices within the Finnish timber industry. One of the main reasons for breeding a tree that conforms closely to this ideotype is to reduce the silvicultural requirement for side pruning of branches by substituting trees that have the ideal growth form without it. Most of the attributes specified by this ideotype would be appropriate to industrial conifer plantations anywhere, but some may be specific to a narrow range of site conditions (e.g., tolerance to heavy snow loading on branches).

This example illustrates another characteristic of the use of ideotypes in the literature: the specifications for different attributes may be given with markedly different levels of precision. Contrast the very detailed specification of branch attributes consistent with the optimum growth form vs. the generalized requirement for "good timber quality." More detailed specifications could have been given for the attributes that constitute "good timber quality" but it simply wasn't very pertinent to do so in an ideotype designed to focus attention mainly on growth form. Ideotypes are flexible creations of a disciplined imagination designed to serve the immediate purposes of the creator. They are often highly expandable and can be refined almost endlessly, as more detailed specifications for desirable attributes become manifest over time. Table 2 gives an example of a highly refined ideotype, which evolved to its present level of precision from a less detailed original formulation in 1975.

Table 2. Attributes of a short-rotation, intensive-culture poplar ideotype for the north central U.S. (Dickmann 1985).

_____________________________________________________________

Growth and phenologv

Fast yearly shoot growth rate (>2m/yr)

Rapid height growth early in the growing season to quickly maximize the leaf area

index and slower growth later in the season to minimize vessel formation Indeterminate growth habit with bud-set just prior to first frosts High shoot/root ratio

Rapid shift of photosynthate flow to roots after bud-set Weak competitor ('crop' ideotype)

Efficient use of soil nutrients

Vigorous stump/stool sprouting, but weak root sprouting; sprouts few in number with strong apical control

Tolerance to pre- and post-emergence herbicides

Resistance to major pathogens (e.g., Septoria musiva, Cytospora chrysosperma, Melampsora medusae, and Marssonina brunnea) and insects (especially defoliators and borers)

Crown and leaves

Large, vertically oriented leaves in the upper crown grading to smaller, horizontally oriented leaves in the lower crown

Leaves with a large photosynthetic rate, specific leaf weight, ratio of net

photosynthesis to dark respiration, and water use efficiency

Leaves, especially in the upper crown, retained until late autumn to maximize the

period of photosynthate production

Relatively few, vigorous but thin lateral branches (to maximize the number of 'phytomeric' units)

High ratio of long (indeterminate) to short (determinate) shoots in the lower crown

stem and wood properties

Excurrent growth habit with straight central stem

Cambium active until late in the growing season

Bark thin, low in stone cell content, and with over 20% fibre content Small pith

Wood low in gelatinous fibres and extractives, with vessel content <20% Specific gravity c. 0.4 g/cm3

Fibre length >0.8 mm

Roots

Vigorous rooting along the entire length of hardwood cuttings

Strongly developed taproot for anchorage and exploitation of soil water and nutrients Dense fibrous root mass in the upper 10 cm of soil Predisposed to infection by endomycorrhizas

Strong sink for photosynthates late in the growing season

_____________________________________________________________

This ideotype, which is largely based on morphological attributes but also includes some physiological and phenological attributes, is intended to fit the following silvicultural system and end-use requirement:

(a) a clone of Populus species or hybrid would be grown from unrooted cuttings on a moist, well-drained, cultivated, fertile site located in the northern Lake States (USA); (b) spacing would be approximately 2x2m; (c) weeds would be controlled for the first two years; (d) the rotation would be six to eight years; (e) mechanical harvesting would be employed, with on-site chipping; (f) whole-tree chips would be pulped for production of linerboard; and (g) stand regeneration would be by stump coppicing (Dickmann 1985).

An ideotype denotes a set of attributes that are appropriate to a particular end use, a particular management system, and a particular set of site conditions. In principle, the purpose for which an ideotype is intended should always be specified with the degree of explicitness exemplified by Dickmann, but unfortunately this is rarely the case. The failure to give explicit and systematic consideration to the purpose for which an ideotype is being proposed can result in being at cross-purposes with what is actually needed.

In industrial forestry the goal of maximum timber production can be virtually taken for granted and there is little harm in failing to make explicit the taken-for-granted assumption that the attributes of the desirable tree should be consistent with the goal of maximum timber production. However, as we shall soon see in our review of the eucalyptus controversy in India, the extension of this taken-for-granted assumption from industrial forestry into the far more diverse and demanding realms of agroforestry and community forestry can give rise to problems.

Phytosociological attributes of trees as a basis for classification of ideotypes

The way trees interact with each other and with other plants may strongly influence their compatibility with the spatial opportunities of different tree planters. Dickmann (1985) provides a succinct overview of how phytosociological attributes have been used by foresters to classify the dominant plantation tree ideotypes:

The forest ecologists' division of species into 'tolerant' and 'intolerant' is nothing more than a reflection of the competitive niche of the species. Donald and Hamblin (1976) defined competitiveness in a manner more appropriate to this discussion. They proposed that plants could be divided into 'isolation,' 'competition' (here referred to as 'dominating'), and 'crop' ideotypes, each with a distinct competitive strategy, and each one appropriate for a particular cultural system.

The 'isolation' ideotype is a free-standing tree which is able to exploit its environment as extensively as possible. It produces a tall, dense and broad crown which displays foliage over a broad area (Cannell 1978), and presumably has a deep and spreading root system. Although the growth and development of an 'isolation' ideotype may be limited by the environment (for instance in the dry savanna), it is nevertheless a strong competitor, able to suppress surrounding vegetation. In terms of tree crops, the 'isolation' ideotype would be appropriate for a fruit, nut or pasture tree.

The 'dominating' [read also 'competition') ideotype tends to overtop neighbouring trees, becoming large at their expense. Stands comprised of this ideotype soon break into crown classes--dominant, codominant, intermediate, and suppressed--and display a wide range of stem diameters. The 'dominating' ideotype would be the model most appropriate for plantations of high-value species, such as Juglans nigra and Quercus robur, where the aim is to produce large-diametered trees yielding high-grade logs. Mortality of suppressed trees is of little consequence in this situation, if the remaining trees are of the highest quality and are growing rapidly. The silvicultural goal may be just a few exceptionally valuable trees per hectare at the end of the rotation

Finally, the 'crop' ideotype is the model appropriate for the fibre or biomass plantation. Individuals of the crop ideotype are not strong competitors, and make efficient use of that portion of the basic site resources to which they have access in a forest community (Cannell 1978). They typically have a dense, narrow crown, with strong apical control, and produce stands with a narrow range of stem diameters. Mortality, and the concomitant loss of accumulated biomass, is not excessive and per-hectare productivity is potentially high. Trees of this ideotype are the forestry equivalents of agricultural field crops, and, like their agronomic counterparts, they require a high level of inputs. Control of weeds during establishment, and the period prior to stand closure, is particularly important. Once established, resources must not become limiting in stands comprised of 'crop' ideotypes, otherwise stagnation may occur and productivity may drop markedly. Therefore, thinning may be necessary in long rotations, because natural thinning would not readily occur, and would not be desirable economically.

It should be noted that the "crop" ideotype referred to here actually denotes a "monocrop" ideotype, and does not at all reflect the attributes that would be required for trees in intercropping systems requiring a high degree of resource sharing (Buck 1986) among the components. In order to deal with this type of situation in agroforestry, Huxley (1985) has expanded the above classification to include what he calls "associative" ideotypes:

In agroforestry, and in systems consisting of mixtures of plant associates, we will mainly be concerned with "associative" ideotypes. These will be selections of either woody perennials, or perennial or seasonal fogs and grasses, each of which can contribute to the fulfillment of the system's objectives, while maximizing environmental resource use by integrating and sharing resources in both space and time. Each kind of plant in such a system could be idealized in terms of a "specification," based on the needs of the system and the technical requirements of available plant types (Huxley 1985).

Extending the range of the ideotype concept

The most frequent users of the ideotype concept have been tree breeders and for this reason ideotypes are usually thought of as referring to characteristics within a particular species. However, in the original sense of the term as a "form which denotes an idea," there is no inherent reason that the ideotype concept cannot be applied before rather than after the species or even the genus has been selected in order to define what kind of tree, or even more broadly what kind of plant, would be appropriate to the needs of a particular application. Once the genus and species were selected, successively more refined ideotypes could then be developed in the conventional intraspecific sense. This proposal may run counter to established convention among forest tree breeders, but what lasting advantage could there be in restricting use of the eminently rational ideotype concept to only the final stages of the plant selection process?

One of the first clear applications of the ideotype concept from the beginning of a process of rational plant selection was given by Felker and Bandurski (1979) in a seminal article that proposes the consideration of leguminous trees as a basis for low-input sustainable agricultural systems. They begin by asking "whether there are plants usable for food and feed, whose physiological, ecological, and morphological characteristics obviate the need for tillage, irrigation and fertilization and provide high yields of protein-rich food?" (Felker and Bandurski 1979). They continue as follows:

To help identify such plants, if in fact they exist, we developed an idiotype [sic] with the following characteristics: 1) growth conditions that cause minimal soil and nutrient loss, 2) little or no need for irrigation, 3) high yield, 4) ability to fix nitrogen and 5) production of large amounts of high quality protein. Next, we have made an assessment of food-producing plants and found that tree legumes have the characteristics of the idiotype....

[Furthermore] ... on the basis of the data currently available it appears that an orchard of leguminous trees with a perennial sod cover would most closely approximate the ideal crop for minimal energy input agriculture.... (Felker and Bandurski 1979)

They then go on to select the mesquite tree (Prosopis) and show how closely the attributes of this genus approximate the ideotype they have specified.

Although it employs the same basic logic as the ideotype concept used in the previous examples, by starting to apply this logic at the earliest stages of the selection sequence-- asking first "what kind of plant?," then "what kind of tree?," then "what kind of leguminous tree?," and finally "which leguminous tree"--Felker and Bandurski illustrate a radically different approach to the planning of technical interventions, one which opens up thinking about the range of options that exist for the use of plant materials to meet clearly defined objectives.

Rather than merely adding precision to the definition of the desirable characteristics of the ideal tree within a predetermined species selection in the context of a taken-for-granted management system, the approach illustrated by Felker and Bandurski is geared toward identifying, through a systematic application of open-ended planning principles, the kind of plant (if any) that would be capable of meeting the performance specifications of a particular application.

"Tree specifications" in ICRAF's D&D methodology

The next step in the application of this kind of "engineering" logic to the design of tree based systems is exemplified by the diagnosis and design (D&D) methodology developed by ICRAF to assist in identifying research and extension priorities in agroforestry (Raintree 1982, 1987a, 1987b; Hoekstra 1985; Rocheleau 1983, 1986; Huxley and Wood 1984; Scherr 1987, 1989). Here, the derivation of "multipurpose tree specifications" is based on the matching of the ideotype to a set of technology design specifications derived not merely from the designer's own idea of what is needed but from a diagnostic analysis of problems and potentials in the existing land use system.

In the D&D methodology the management objectives of the existing land user are taken as the starting point for the diagnostic analysis. The performance of the relevant output subsystems is then assessed to detect varying degrees of success or failure in meeting the land user's objectives. A significant mismatch between actual and desired performance then invokes a trouble-shooting procedure to identify the causes of poor performance.

An analysis of the key constraints within the network of causal factors behind the problem is then used to identify functional intervention points at which the performance of the system could be significantly improved. The diagnostic phase of the process concludes with the writing of general "system specifications" that define the kind of technology that would be capable of effecting the desired functional intervention in a manner most appropriate to the diagnosed land use system. These specifications are then carried forward into the design phase.

The design phase begins by reviewing the range of technology options that could conceivably meet the system specifications. In keeping with ICRAF's policy of attempting to be an "honest broker" in agroforestry (Steppler 1981), both agroforestry and non-agroforestry options are considered at this stage. This "brainstorming" step is then followed by a more critical analysis leading to a short list of the most promising "candidate technologies." At this stage in the planning process, the agroforestry design concepts are usually singled out for further attention. Detailed "technology specifications" are written by the multidisciplinary D&D team to record the design concept that will guide the initial research and development of prototype technologies.

Foremost among the design specifications set down at this point are the "MPT specifications," which provide criteria for multipurpose tree species selection and germplasm screening. They define the attributes of the ideal multipurpose tree for the envisaged agroforestry intervention. Because trees are limited by their adaptive strategies to certain "allowable" combinations of attributes (Huxley and Wood 1984), it may require more than one tree to meet all of the specifications in an idealized design. An example of diagnostically derived ideotype specifications for an agroforestry application is given in Table 3.

Table 3. Tree/shrub specifications for hedgerow intercropping by small farmers on sloping lands in Malawi (Minae 1989).

See also indicative specifications for other tree growing practices in Appendix D.

_____________________________________________________________

· compatibility with climatic and edaphic parameters of the site

· fast-growing

· easy to establish

· N-fixing (or other nutrient cycling ability, e.g., mycorrhizal association)

· deep-rooting habit (for drought tolerance, recycling of leached nutrients and minimal surface root competition with crops)

· ability to coppice vigorously and produce large quantity of herbaceous material under heavy pruning (for green manure/mulch)

· ability to provide by-products of fuelwood and/or fodder under appropriate management

_____________________________________________________________

The D&D approach is similar in intent to other "microplanning" approaches used in community forestry (Hardcastle 1987, Banerjee 1988) but it is somewhat unique in its specific ability to define the socioeconomically relevant attributes of the required trees prior to species selection. The ICRAF approach provides a multilevel framework for linking "tree specifications" with higher order "technology specifications" with still higher order "system specifications" through a client-oriented diagnostic approach. A related development, which only came to the author's attention as this manuscript was going to press, is the work on farmer defined multipurpose tree ideotypes supported by the Forestry/Fuelwood Research and Development Project within the Multipurpose Tree Species Research Network in Asia (MacDicken and Bhumibhamon 1990, MacDicken and Mehl 1990, Chuntanaparb and Ranganathan 1990).

Burley and Wood (forthcoming) underscore the special importance of correct ideotype definitions for multipurpose trees. After emphasizing that good "proposals for agroforestry designs arise from the diagnostic and design phases of work... and result in a description of an 'ideal tree' or trees for meeting the production and service roles required by the system," they write:

Of course, no single species or individual tree can exhibit all the traits that may be desirable for an agroforestry application. Indeed some of these traits may be mutually exclusive, such as fruit and wood production, as these may often represent competing sinks for photosynthates. Any particular set of morphological and physiological characteristics displayed by a plant can be related to its past evolutionary history. The very fact that MPTS are needed for multiple objectives emphasizes the need for precise definition of the desired ideotype. As the plants are often to be grown in association with plants both of the same kind, and also with others of different morphology and phenological behaviour, a particularly wide range of attributes and tree characteristics has to be considered in determining appropriate ideotypes (Burley and Wood, forthcoming).

As an aid to tree selection for specific functions Carlowitz (1986) has provided the following checklist of multipurpose tree (MPT) attributes:

Table 4. Attributes of multipurpose trees in relation to possible production and service functions.

Tree management as an intervening variable

Clearly, the attributes that make a tree suitable for one system of management might make it unsuitable for another (Cannell 1983, Huxley 1985), but management of the tree biomass might also be used to modify the expression of biophysical attributes and thus substitute for genetics.

The examples given above of temperate forest plantation tree ideotypes define trees that are genetically equipped to accomplish their desired purposes with minimal management inputs. Indeed, the substitution of genetically engineered characteristics for costly silvicultural operations may be an explicit aim of forest tree breeding programmes. In agroforestry, however, as the example of the hedgerow intercropping ideotype and the above list of MPT attributes might suggest, the premium may often be placed on attributes that enable the trees not to substitute for management operations, but to respond in a desirable way to prescribed management operations within an agroforestry cropping system. Key elements of such a design might include, in addition to the appropriate tree ideotype, the spatial arrangement of the trees and other plant components, as well as the management operations to be carried out on the entire complex.

Given the choice of two ways of achieving the same morphological or performance characteristics, a) "automatically" by genotype selection or b) "deliberately" by phenotype management, there are situations in which one might well have a strong preference for the latter--typically, when the management operation yields a useful "by-product." In the hedgerow intercropping example, less vigorous growth of pruned branches would make it easier to maintain a compact "hedge" shape and thus control competition with the adjacent crops, but the main reason for pruning is to harvest the green manure, fodder, or fuelwood regrowth (the proportion of each depending on the nature of the pruning regime). Since the by-products of the management operation are in themselves of value, it is appropriate to select trees capable of vigorous regrowth and use labour inputs to maintain the hedge shape and control the flow of by-products.

Management inputs (labour costs) that might be regarded as an economic liability in a single purpose management system, might actually be the key to increased economic yields in a multipurpose management system. Management operations are the means of extracting additional benefits from trees in an agroforestry cropping system but there is usually a trade-off with labour inputs. Generally speaking, the greater the number of management operations applied to extract additional benefits, the greater the labour requirement of the tree growing technology.

Management flexibility can be a top priority in such systems in order to allow the fanner to respond to seasonally changing needs. For example, depending on how the year is shaping up, a farmer might decide to alter the proportion of mulch or fodder or stickwood obtained from a hedgerow intercropping system by changing the frequency or height of cutting.

Let us now return to the question that was raised and partially answered at the beginning of this chapter, "Do trees have socioeconomic attributes?" We can now say quite clearly that trees have inherent biophysical attributes that make them more or less compatible with different potential users but that the expression of these attributes and their perception by different users are conditioned by the way in which the trees are managed. In other words, management is an intervening variable between the biophysical attributes of trees and the socioeconomic expression of those attributes.

We shall also see in the next chapter that trees may play powerful roles as symbols of concerns that have little or nothing to do with their biophysical attributes but that add yet another layer of socioeconomic meaning. To the extent that they all present different opportunities to different users, these different types of attributes--intrinsic, intrinsic-modified-by-management, and ascribed--all refer to what we might usefully call the "socioeconomic attributes" of trees.