II. Ecological and social research

Experimental science and development: A re-evaluation of the environmental crisis hypothesis, 1939-1960


The African environmental crisis hypothesis (based on equilibrium ecological model) has been popularized by colonial and post-independent development and land-use policies.1 The crisis hypothesis is closely associated with the general desiccation hypothesis discussed in Chapter 2 and historical patterns of land use discussed in Chapter 4. Although specific processes that defined the environmental crisis have not been clearly described—perhaps mainly due to lack of verification—the hypothesis has developed a life of its own.2

The current chapter presents a meta-analysis of published proxy environmental indicators (sediment production, crop yields, stream discharge, range production and livestock production performances, reseeding and bush clearing) from agronomic and range science grazing experiments to investigate if the outcomes of experiments may be interpreted according to the equilibrium or the alternative disequilibrium hypothesis (the following section discusses these two hypotheses). The experiments are those published in issues of the East African Agricultural arid Forestry Journal between the 1930s and 1960s. We selected experimental reports that met the following conditions: (1) the experiments were scientifically designed (i.e., with replications of treatments and controls);’ (2) the experiments simulated different land-use intensifications;4 (3) the experiments were conducted for at least two years.

We specifically selected reports in which the data had been clearly displayed according to treatments, controls and time—which made it easy to summarize for visual comparisons. We used the interpretations of the authors and our own to re-appraise and verify the contrasting ecological hypotheses (equilibrium vs. disequilibrium). We present our findings in the following sections: (1) ecological theoretical framework—describing the features and assumptions of the equilibrium and disequilibrium hypotheses; (2) rationalizing experimental protocols; (3) agronomic research; (4) range science research; and (5) testing the environmental crisis hypothesis from the perspectives of the opposing ecological models.

Ecological theoretical framework

The history of imperial science in Africa has been framed around environmental degradation—associated with indigenous land use, particularly in arid and semi-arid environments. As introduced in Chapter 1, the ecological basis for considering land degradation is the succession model (also called the equilibrium model). This model assumes that environmental events are self-regulating unless the processes are disrupted by disturbances often associated with land-use activities. It proposes that land-use intensification serves as a trigger to prompt environmental indicators to change unless management initiatives intervene. Land-use intensifications include the perceived overcultivation of agricultural lands and overgrazing of grazing lands. The model further supposes that rangelands—including those in Africa—have fixed carrying capacities which provide a ceiling for allowable stocking rates, beyond which the range degrades.

While imperial science has focused on the proximate causes of land degradation (i.e., human actions), the ultimate causes (e.g., rainfall variability) that trigger different processes described by the disequilibrium model have been overlooked.’ We have mentioned before that rainfall variability would influence land-use changes in arid and semi-arid regions, including those in Africa.6 Thus, when during later years, the thinking about land degradation shifted from proximate to ultimate causes, this represented a paradigm shift in considering ecological processes that influence land degradation.' According to this alternative disequilibrium model—which clearly includes the functioning of the drylands—carrying capacity of the rangelands are highly variable in space and time in accordance with rainfall variability. The disequilibrium model is scale-dependent—this implies that land-use impacts at the scale of land units (i.e., local landscapes) will respond differently from those at geographical scales. Putting it another way, it would be inappropriate (without adjustments) to apply scientific results from research stations that pronounce problems of land degradation at restricted scales to indigenous land use on a larger geographical scale.

Further, the equilibrium model advocates rest and rotational grazing as alternatives to continuous rangeland grazing.8 Fallow periods, reseeding and bush clearing were expected to improve rangeland production, while continuous grazing by livestock was expected to aggravate the problem of range restoration. In the context of these opposing views, we now rationalize a selection of proxy environmental indicators in selected experimental protocols to test the environmental crisis hypothesis.

Rationalizing experimental protocols

Testing the equilibrium model (which precipitated the environmental crisis hypothesis) requires an upfront definition of proxy environmental indicators of land-use intensification that have been investigated across East

Africa. These include sediment production, soil moisture, soil fertility, storm discharge, stocking rates, range production and conditions, and trends.9 Land-use intensification may be described broadly as the expansion of agronomic and rangeland production, with associated changes in land- use gradients and the performance of corresponding proxy environmental indicators.10 The response of the proxy indicators is inferred from their direction of change (i.e., increasing, decreasing or no response) as a result of land-use intensification. We have framed the relationships between land- use intensification and changes in the proxy environmental indicators using a schematic scenario model" (Figure 5.1). Each scenario provides an ecological explanation (either equilibrium or disequilibrium) in relation to changing land-use intensification. Considering that different proxy environmental indicators behave differently, each scenario might represent varieties of land-use intensification independently—that is, an increase in one type of indicator might not necessarily imply a decrease in another and vice versa.

In Figure 5.1, scenarios A and D might imitate the presumed environmental crisis hypothesis, while scenario C represents the alternative disequilibrium hypothesis. Scenario A corresponds with increasing trends of the proxy indicators with increasing land-use intensification (e.g., increasing sediment production). Conversely, scenario D shows declining trends of the proxy indicators (e.g., diminishing soil fertility) in response to land-use intensification. Scenario В is controlled by factors other than land-use

Scenarios of hypothetical intensities of land use and corresponding proxy environmental indicators

Figure 5.1 Scenarios of hypothetical intensities of land use and corresponding proxy environmental indicators.


The scheme shows the predictions of equilibrium (environmental crisis hypothesis) (A and D) and (C) disequilibrium hypothesis. Line В corresponds with an independent gradient.

intensification (i.e., it is independent). In scenario B, the lack of response does not imply that the proxy indicators are not suitable for decision making in cases where environmental variables are appropriate for measuring changes. Scenario C (disequilibrium) corresponds with the ‘intermediate disturbance’ hypothesis12 that suggests that potential responses reach a maximum at the level of medium land-use intensification. In our representation it might also suggest spatial and temporal variability of the indicators that would be necessary to calibrate to the scales of management.13

Colonial officials perceived that intensification of crop cultivation and overstocking of the rangelands would cause spiraling degradation due to increased soil erosion (Scenario A). In both cases, the proxy indicators respond in a feedback, namely that increased land-use intensification is followed by positive (scenario A) or negative (scenario D) changes in the indicators. In the view of imperial science, the African herders managing multi-species livestock would allow them to graze the vegetation closely. Accordingly, it was perceived that

if not systematically controlled [the livestock] will crop an area so closely that the grass is not even permitted to come to flower and sheep will even pull up and eat the roots of many kinds of grass, whilst the goats will do untold damage to the trees and shrubs.14

However, two parallel processes might be operating in the grazing system. In the first process, until the level of medium land-use intensity (scenario C), grazing promotes rangeland productivity.11 For scenario C, a decrease in the indicator with higher land-use intensity assumes that animals would continue to nibble at individual plants. Yet, in practice this rarely happens in African savannas where plants are highly adapted to herbivory. Therefore, in the second process, perennial plants use defensive mechanisms (such as thorns and spines) to protect the stubble biomass.16 Hence, the grazers adapt their behavior by moving from one plant to another. Scenario D is rarely achieved in the field. The final outcome is influenced by rainfall variability rather than by grazing. It would therefore be imprudent to ignore factors such as droughts that might depress plant biomass, or sediment production from a watershed. We now apply some of these scenarios to the topic of agronomic research.

< Prev   CONTENTS   Source   Next >