Previous Page Table of Contents Next Page

Chapter 5
Soil erosion and loss of productivity

The effect of soil erosion can be measured in different ways according to the kind of damage suffered. In the model (Figure 2.1), the estimate is based on short-term losses in crop production due to erosion of fertile topsoil, and long-term losses in land productivity due to truncation of the soil profile and consequent reduction of available water. No account is taken at this stage in the model development of possible damage to lowlands by flooding and silt deposition, or of the possible benefit from the deposition of fertile silt on alluvial plains, or increase in workability constraints due to changes in terrain characteristics.

In the model, permissible slopes for various land uses under different levels of inputs circumstances have been defined as model variables, and these are given in Table 5.1. The critical slope values in the slope-land use association screen define the upper slope limits to cultivation, and they may be modified as appropriate.

Further, the model takes into account the loss in crop production by soil erosion through:

  1. the removal of topsoil which, in many soils, is the source of most or all the nutrient fertility; and

  2. reducing the overall depth of the soil profile so that eventually the soil water holding capacity and foothold capacity are reduced to a point where it limits yields.

An acceptable rate of soil erosion is considered to be one that over a specified number of years (e.g. 25, 50 or 100):

  1. does not result in a crop yield reduction of more than a specified amount due to loss of topsoil; and

  2. does not result in more than a specified proportion of land being downgraded to a lower class of agricultural suitability due to soil depth reduction.

These two criteria are not interdependent, so that acceptable rate of soil loss is taken as the lower of the two alternatives. The model therefore provides a framework for assessing tolerable soil loss, based on its likely impact on crop yields and the future availability of cultivable land.

Slope-cultivation association screen

Land utilization typeLevel inputs
Dryland crops without soil conservation measures< 30%< 30%< 16%
Dryland crops with soil conservation measures< 30%< 30%< 30%
Wetland crops without soil conservation measures< 5%< 5%< 2%
Wetland crops with soil conservation measures1< 30%< 30%< 30%
Coffee, tea, fuelwood and pasture with and without soil conservation measures< 45%< 45%< 45%

1 For wetland crops, terracing is required.

Regeneration capacity of topsoil (mm/year) by length of growing period (LGP) and thermal zone

Thermal Zone
< 750.
75 – 1191.
180 – 2691.
> 2702.

Derived from Hammer (1981).

The soil erosion and productivity model (Figure 2.1) is linked to crop, livestock and fuelwood productivity models which provide the assessments of land suitabilities and the associated yield potentials for the estimation of tolerable soil loss.

5.1 Effect of Topsoil Loss on Productivity

Soils differ in their susceptibility to loss of productivity as the topsoil is eroded. The differences are related to the depth of the topsoil and the amount of nutrient fertility or presence of unfavourable conditions in the subsoil.

Loss of productivity due to topsoil loss can be largely compensated by the use of manure and fertilizer, and low rates of soil erosion are compensated to some extent by the formation of new topsoil. The rate of topsoil formation can vary from < 0.25 mm/year in dry and cold environments to > 1.5 mm/year in humid and warm environments (Hammer 1981; Hudson 1981). Topsoil formation at the rate of 1 mm/year is equivalent to an annual addition of 12 t/ha. Therefore, the rate of topsoil formation has been considered as a factor in the model in assessing loss of productivity and tolerable soil losses. Regeneration capacities of soils used in the model in calculating net loss of topsoil are given in Table 5.2 by moisture and thermal regimes.

Ranking of soils (Kenya Soil Survey) according to productivity loss per unit of topsoil

Most susceptibleIntermediate susceptibleLeast susceptible
Acrisols, except Humic AcrisolsArenosolsChernozems
Ferralic CambisolsCambisols, except Ferralic CambisolsFluvisols
Ferralsols, exept humic FerralsolsGleysolsHistosols
Ironstone soilsGreyzemsHumic Andosols
LithosolsHumic AcrisolsMollic Andosols
PlanosolsHumic FerralsolsVertisols
 Vitric Andosols 

Relationships between topsoil loss and yield loss

Soil susceptibility rankingLevels of inputsEquation
Least susceptibleLowY = 1.0 X
 IntermediateY = 0.6 X
 HighY = 0.2 X
Intermediate susceptibleLowY = 2.0 X
 IntermediateY = 1.2 X
 HighY = 0.4 X
Most susceptibleLowY = 7.0 X
 IntermediateY = 5.0 X
 HighY = 3.0 X

Y = productivity loss in percent;

X = topsoil loss in cm.

Based on experimental evidence (Stallings 1957; Barr 1957; Lal 1976a, 1976b, 1976c; Higgins and Kassam 1981) and analytical data from Kenya Soil Survey (KSS 1975, 1976, 1982b), soil units of the Exploratory Soil Map of Kenya have been classified according to their susceptibility to productivity loss with loss of topsoil, and on the presence of other unfavourable subsoil conditions (Table 5.3). These rankings of susceptibility of the soils are related to actual yield losses, by inputs level, through a set of linear equations given in Table 5.4. The reduced impact of topsoil loss under intermediate and high levels of inputs is due to the compensating effect of fertilizers at their normal rates of use. It is assumed that the benefit of fertilizers is less on the more susceptible soils because of their more unfavourable subsoil conditions.

The tolerable loss rate, for a given soil unit and specified amount and time scale of yield reduction, is calculated in the model as follows:

TL = {(Ra/Rm × 100 B × Dt) + 3T} / T(5.1)

where:TL=tolerable loss rate (t ha-1 year-1)
 Ra=acceptable yield reduction (%)
 Rm=yield reduction (%) at the given inputs level when the effective topsoil is all lost
 B=bulk density of the soil (g/cm3)
 Dt=depth of effective topsoil (cm)
 T=time (years) over which yield reduction is acceptable.

5.2 Effect of Soil Depth Reduction on Productivity

The rate of soil formation by rock weathering is extremely slow, up to 0.025 mm/year on volcanic rocks in humid areas, and < 0.01 mm/year on basement complex rocks in semi-arid areas (Dunne, Dietrich and Brunego 1978). At the highest rate quoted by Dunne et al. (1978), it would take 4,000 years to produce 10 cm of soil. Therefore, the rate at which the soil profile is deepened by rock weathering has not been considered as a factor in the model in assessing tolerable soil losses.

The estimation of the effect of soil depth reduction is based on the assumption that there is no significant loss of productivity until the soil becomes so shallow that shortage of moisture becomes a limiting factor. The critical depth varies according to crop and the climate. Once this critical depth is reached, productivity loss is linear until the soil becomes too shallow to produce any crop at all (Wiggins and Palma 1980). The critical points can be equated with land suitability class limits as follows (where depth is the limiting factor):

VS/S:Soil water becomes limiting and there is at least 20% decrease in yield potential
S/MS:Soil water becomes limiting and there is at least 40% decrease in yield potential
MS/mS:Soil water becomes limiting and there is at least 60% decrease in yield potential
mS/N:Soil water becomes limiting and there is at least 80% decrease in yield potential

The land suitability classes VS (very suitable), S - (suitable), MS (moderately suitable), mS (marginally suitability) and N (not suitable) correspond to yield levels of >80%, 60–80%, 40–60%, 20–40% and < 20% of maximum attainable yield respectively.

If erosion takes place uniformly on soils of varying depth, the end result will be that some soils that had been marginally deep enough will become non-productive while others will become marginal. If the range of soil depths is known, the tolerable amount of soil loss can be gauged in terms of the amount of land that can be permitted to be lost to production.

In order to calculate tolerable soil losses, soil depth reduction is measured in terms of the proportion of the soils in a specified area that have become shallower than a given depth, as a result of erosion. The soils of the mapping units of the Exploratory Soil Map of Kenya (KSS 1982a) are assigned to 5 depth classes: shallow, < 50 cm; moderately deep, 50–80 cm; deep, 80–120 cm; very deep, 120–180 cm; extremely deep, > 180 cm. The rate of soil loss is related to the proportion of land whose soil has become shallower than a specified depth, by the following equations:

  1. Proportion (P, percent) of land downgraded to at least the next depth class:

    P = (SL × T) / (B × Dr)(5.2)

    where:SL=soil loss (t ha-1year1)
     T=time (years)
     B=bulk density of the soil (g/cm3)
     Dr=depth range of the soil class (cm).
  2. Proportion (P, percent) of land downgraded by more than one depth class:

    D2=difference (cm) between the lower limit of the depth class and the upper limit of the shallower class to which the land is downgraded
    Dr=depth range of the soil class (cm). Table 5.5 shows the proportions of land downgraded from given depth classes to shallower classes as a result of soil erosion at different rates over a 100 years period. The values in Table 5.5 are based on equations 5.2 and 5.3, and assume that soil depths are evenly distributed over the range in each depth class.

If a tolerable soil loss was set to allow 10% of each depth class to be downgraded by one class over 100 year period, this would give the following soil loss rates for each depth class (assuming 25 cm is the minimum soil depth that would allow crop production):

Shallow (to 25 cm)- 3 t ha -1 year -1
Moderately deep- 3.6
Deep- 4.8
Very deep- 7.2
Extremely deep- 26.8

5.3 Assessment of Tolerable Soil Loss on a Combined Basis of Topsoil Loss and Soil Depth Reduction

Criteria for estimating soil loss tolerance are set according to the amount of yield loss that can be tolerated, or the proportion of the land that can be permitted to become shallower than a specified depth, over a specified time. The two basis for soil loss estimation do not interact, so when used in combination the tolerable soil loss would be the lower of the estimates. An example of the soil losses that would give either a 50% yield reduction or soil depth reduction resulting in downgrading of 10% of each depth class, over a period of 100 years, is given in Table 5.6.

The proportion of land downgraded from given depth classes to shallower depth classes or to bedrock as a result of soil erosion at different rates over a 100 year period

Soil depth class and change
Amount of land downgraded (% of class)
at erosion rates (t/ha) of:
From shallow (0–50)        
to bedrock (0)8174283100   
From moderately deep (50–80)        
to shallow (0–50)142870100    
to bedrock (0)000042100  
From deep (80–120)        
to moderately deep (50–80)102152100    
to shallow (0–50)0002981100  
to bedrock (0)   008100 
From very deep (120–180)        
to deep (80–120)7143570100   
to moderately deep (50–80)00033872100 
to shallow (0–50)   0022100 
to bedrock (0)     078100
From extremely deep (200–400)24919283876100
to very deep (120–180)000011148100
to deep (80–120)    0030100
to moderately deep (50–80)     01792
to shallow (0–50)      070
to bedrock        

Tolerable rates of topsoil loss (t ha-1 year-1) to give not more than 10% loss of land from a given depth class and not more than 50% crop yield reduction at low input level over a 100-year period

Soil depth classSusceptibility to yield loss of topsoil
Moderately deep3.63.63.6
Very deep7.27.27.2
Extremely deep12.026.426.4

1 Assuming a minimum depth of 25 cm for crop production.

Previous Page Top of Page Next Page