(introduction...)

Investigation of higher order linkages in two dimensions can be carried out by using directional diagrams called networks. Networks, although widely discussed in the EIA literature, have not been used as extensively as matrices and simple checklists. Networks were essentially developed to explicitly consider the secondary, tertiary, and higher order impacts that can arise from an initial impact. Here, any effect on the biophysical and socio-economic environments that arises from a cause directly related to the project activities is termed a first-order or primary impact. The secondary impacts are those affecting the biophysical and socio-economic environments which arise from an action, but which are not initiated directly by that action. Presentation matrices can only clearly show the primary or first-order impacts within any particular activity-component framework.

Figure 4.3 symbolically traces the secondary impacts which could arise from the dredging and filling of an estuarine mud-flat (supposing that the proposed project was the construction of a marina). The top row of rectangles shows a set of environmental components linked by various dependencies (i.e., second row of boxes). The way in which the primary impact of dredging and filling the mud-flats affects all the components is shown as a progression of causes and effects (viz. in rows three and four). Note that the commercial facilities component of the socio-economic system has a fourth-order dependency on salt marsh plants. Similarly, the action of dredging and filling the mud-flats has a fifth-order impact upon commercial facilities.

Various authors have used different terms to describe secondary impacts, but most are compatible with the above definition. Jain and Urban (1975), however, use the term "higher order impacts'' to mean secondary impacts as given above, while reserving the terms "indirect'' or "secondary'' to cover impacts resulting from an induced action.

Box 4.4 presents a checklist of second-order impacts arising out of the use of pesticides in an irrigation project.

The network technique developed by Sorensen is probably the best-known approach for investigating higher order impacts. The objective of the network approach is to display, in an easily understood format, the intermediate links between a project and its ultimate impacts. This type of network includes the identification of probable importance of temporal effects as well as a list of data requirements. Complexity increases as higher order impacts are considered, and, as a result, the Sorensen network is restricted to third and lower order impacts. Figure 4.4 shows a portion of a network devised by Sorensen to display the possible consequences of various forms of land use for a section of Californian coastline. Three options for residential development are related to four primary impacts, and cause-effect linkages are developed for each identified primary impact. The diagram also takes notice of feasible mitigatory measures. This type of network includes the identification of probable importance of temporal effects as well as a list of data requirements.

Figure 4.3 Symbolic representation of the secondary impacts which could arise from the dredging and filling of on estuarine mud-flat. The relationship between environmental components, their interdependencies, and the causes and effects of secondary impacts are illustrated

Source: J. B. Shopley, M. Sowman, R. J. Fuggle, Environmental Management. 31, pp. 197-213, 1990.

Box 4.4 Example of secondary order impacts: Use of pesticides in irrigation projects

• Use of pesticides/insecticides/herbicides etc., in excessive quantities and the persistent nature of pesticides can contaminate the soil. • Pesticide-contaminated soil contaminates the groundwater through leaching. • Contaminated groundwater affects human health through ingestion. • Contaminated soil reduces crop yield. • Contaminated soil pollutes surface water through run off from the soil. • Contaminated surface water affects aquatic biota through toxic contamination. • Contaminated surface water decreases yield of fish. • Use of pesticides/insecticides/herbicides, etc., often kill the natural predators. These predators eat the pests and hence are useful in maintaining the crop yield. Thus killing of natural predators by pesticides increases the population of pests and thus reduces yields. • Use of pesticides/insecticides/herbicides etc., affects crops through deposition

Methods are available for translating networks into mathematical models. These methods structure the relationships implied in qualitative simulations. Two common quantitative simulation models are GSIM and KSIM. In GSIM the verbal expression of relationships is made explicit at the simplest level: "if A increases then B will decrease (or increase or be unaffected)''. In KSIM models the relative magnitude of the relationship must be specified: "if A doubles then B will decrease by 25%''. The KSIM model at an extreme level approaches quantitative simulation modelling.