Systems approaches to water management research

https://doi.org/10.1016/S0378-3774(98)00097-3Get rights and content

Abstract

This paper provides an introduction to systems approaches to water management research. Common concepts in systems thinking are defined and the concepts of level, system boundaries and emergent properties are described. Differences between hard- and soft-systems approaches are presented with examples. Finally, the spectrum of approaches to water management research are discussed in terms of their level and degree of holism and in the context of the U.K. Department for International Development's Renewable Natural Resources Research Strategy. A greater emphasis on more subjective and holistic approaches is recommended in order to define more clearly the researchable constraints in water management.

Introduction

The U.K. (DFID) Renewable Natural Resources Research Strategy (RNRRS) `seeks to achieve uptake of research results and impact on livelihoods by supporting systems-based research projects that adopt participatory approaches for identifying farmers' needs and systems constraints' (Walker, pers. comm). The term `systems-based research' is used widely and often imprecisely by researchers and research managers alike. Systems research is, in our view, characterised by having almost as many approaches as there are practitioners! This paper will therefore attempt to define some of the key concepts of current systems thinking.

A system may be considered as interconnected set of components which `behaves as a whole in response to stimuli to any part' (Spedding, 1988). The concept of a system should be familiar to anyone who has worked in irrigation. Hydrologists work with `hydrological systems', engineers with `hydraulic systems' and agriculturalists with `farming systems'.

In all natural or agricultural systems there exists a hierarchy of levels. Each system is, in fact, a component of a super-system and itself comprised of a collection of sub-systems.

For example, Dent (1975)identified four levels of agricultural systems research:

  • biochemical and physical systems;

  • plant and animal systems;

  • farm business systems; and

  • national and international systems.

More recently, Izac and Swift (1994)suggested five levels of agricultural systems:
  • cropping systems;

  • farming systems;

  • village/catchment systems;

  • regional systems; and

  • supra-regional systems.

Interestingly, Izac and Swift do not consider levels lower than the cropping systems (and thus have no equivalent for Dent's biochemical and physical systems) nor higher than the supra-regional system. The highest level in the hierarchy, the supra-regional system, corresponds with the concept of ecoregions that now forms a focus for debate on the future of international agricultural research programmes.

Sub-systems may be considered as `holons' in that they are `whole' in themselves, and at the same time `parts' of larger wholes (Uphoff, 1996). Certain aspects of the behaviour of a sub-systems may be confined entirely to that sub-system, yet systems at any level are constrained and controlled by the levels above and below. For example, individual farmers determine what will happen within their own farming system but must work within the social, political, technical, economic and environmental constraints imposed by the village or catchment system. At the cropping system level, the physico-chemical factors regulating plant production are integrated, whereas human interventions in the form of labour, machinery and management can be regarded as external (Izac and Swift, 1994).

At the farming system level, biological, economic and social considerations are integrated as farmers make resource allocation decisions between different systems within their control. These decisions relate to their overall objectives which are likely to include food security as well as profit generation. The decisions made by individual farmers or farm families will be constrained by their access to resources and may result in some parts of the land degrading relative to the rest. An obvious example is shifting agriculture where, for short periods, resources are preferentially used in one plot of land whilst they accumulate slowly over time in other areas. Water harvesting is another case where resources are transferred from a wider area to provide an increased likelihood of successful crop production in a limited area.

There are no pre-determined boundaries to systems. The boundaries of a system are defined by the observer for a particular purpose. In some instances it may be appropriate to define physical, or spatial boundaries. For example, when studying an irrigation project, the natural boundary for hydrological studies may be the command area, however, the socio-economic systems has no physical boundary but includes all the stakeholders, many of whom may reside outside of the command area. Within the human context, individual farm enterprises interact at the village or community level with other farming and non-farming families sharing resources such as drinking water, communal grazing, and firewood.

A closed system is one in which the system boundaries as well defined and `impermeable' that is, there is no transfer of matter, energy or information outside the system as defined. In order to make analysis easier and apply the law of conservation of matter, it is common to assume that systems are closed. For example, Seckler (1996)argues that as water supplies become more limiting, so water basins tend towards `closed' systems. He argues that water `losses' from one part of the system are `gains' to another. Drainage water, for example, is frequently returned to the system and utilised by downstream users and water is only lost from the system if its quality is affected to the extent that it can no longer be useful.

Many natural and social systems, however, are not closed. They are open systems, in that their boundaries are ambiguous, expanding or permeable.

Fundamental to the holistic approach is the concept of the `whole being greater than the sum of the parts' because of the interactions and inter-relationships between the components (Rountree, 1977). Thus properties emerge from an assembly of sub-systems which are not possessed by the constituent sub-systems independently. Whole systems cannot therefore be properly studied by looking at each component in isolation since the parts interact so that the system exhibits distinct and identifiable `behaviour' and responds to changes in external conditions (Spedding, 1988). In open systems Uphoff (1996)argues that the `zero-sum' way of thinking is inappropriate and excludes `positive-sum', multiple satisfactions resulting from the emergent properties of the system.

An example of an emergent property in an irrigation scheme might be the development of small service industries which arise due to the increased prosperity of those involved in the scheme. Without the technological irrigation system, the natural environment and the social system interacting together such additional prosperity would not be possible. Similarly, better education and greater `well being' tend to result from wealth created in successful schemes as people's expectations change from personal survival through family responsibilities to wider social obligations.

Not all emergent properties need be positive. A negative example is Schistosomiasis which only becomes apparent when the hydrological, social and natural systems interact to provide the right conditions to complete the life cycle.

The concepts of system levels, boundaries and emergent properties introduced above provide a framework for the following review of `hard' and `soft' systems approaches which are then related to issues of better water management.

Section snippets

Systems approaches to research

Systems approaches have been proposed for many as not only being applicable to all scientific disciplines but also as a useful method of integrating disciplines (Boulding, 1956). Many of the approaches adopted by researchers tend to strive for objectivity and can thus be classified as hard systems (see Fig. 1).

Approaches to research can be viewed in terms of two axes of perspective – the degree of `reductionism' and degree of `subjectivity'. The reductionist approach is to break down complex

Discussion

It is useful to consider research in the context of the Hawkesbury hierarchy presented in Table 1. The hierarchy covers the whole range of research and necessarily involves greater inter-disciplinarity and also greater uncertainty as activities move up from basic through applied to hard and finally soft systems research. Applied and basic research are susceptible to a reductionist approach whilst hard and soft systems research require a more holistic view.

Table 1 attempts to identify those

Acknowledgements

The authors would like to thank Richard Carter and other members of the Natural Resources Management Department at Silsoe, and Susan Carr of the Open University for the constructive comments and contributions to this paper.

References (32)

  • Bawden, R.J., Ison, R.L., Macadam, R.D., Packham, R.G., Valentine, I., 1985. A research paradigm for systems...
  • Bayliss-Smith, T.P., 1982. The Ecology of Agricultural Systems. Cambridge University Press,...
  • K.E. Boulding

    General systems theory – the skeleton of science and management

    Manage. Sci.

    (1956)
  • Chambers, R., Pacey, A., Thrupp, L.A. (Eds.), 1989. Farmer First: Farmer Innovation and Agricultural Research....
  • Checkland, P.B., 1981. Systems Thinking, Systems Practices. Wiley, Chichester,...
  • Conway, G.R., 1983. Agroecosystem analysis. Imperial College Centre for Environmental Technology Series E. Imperial...
  • Cited by (0)

    View full text