Introduction

This article contains a presentation of a soil erosion and water management trial site located in the Iringa Region of Tanzania. Special emphasis is given to the performance of Vetiver grass. The aim of the experiments is to evaluate the effectiveness of different low input soil erosion control measures and impact of these measures on the plant nutrient - and water balance. The experiments are fully financed by the Danish International Development Agency (DANIDA) and carried out by the Chemistry Department and Department of Agricultural Science at the Royal Veterinary and Agricultural University, Denmark. Locally, there is close co-operation between the research project, the HIMA/DANIDA-Iringa project (presented in Vetiver Newsletter No. 16) and the Soil Science Dept. at the Sokoine University of Agriculture in Tanzania.

The three soil erosion control measures tested in this study are either recommended by the HIMA-Iringa project and/or practised by some local farmers. Although Vetiver grass in numerous experiments has proven to have almost ideal properties when used to combat water caused soil erosion, no controlled experiments on the effectiveness of Vetiver grass and magnitude of soil erosion in the Iringa Region have been carried out. Data from 3 growing seasons has been collected so far starting with the 1995/96 season. Since this is an ongoing research project and selected data will be published in technical/scientific journals, only few figures are presented here.

The experiments

The trial site is located in Kilolo Division, Iringa District, Iringa Region on a 29% slope with a SSW aspect. The climate is moderately cool tropical and subhumid. The unimodal rainfall pattern is often broken by a short dry spell in January/February. The start of the rain season is marked by thunderstorms yielding high intensity downpours with a high degree of spatial variability. The highest temperatures occur in November/December and the lowest in June/July, when temperatures can drop below 0(C.

Average annual rainfall in the area is 1100 mm and the annual average potential evapotranspiration about 970 mm.

In the area, characterised by steeply dissected convex slopes, maize sometimes intercropped with beans is the dominating crop on well-drained soils.

The nature of soils in the area is determined by the parent material, climate and the extent of weathering. Located in the subhumid and cool tropical climate of the Iringa Highlands and developed on a mosaic of acidic pre-Cambrian metamorphic and plutonic rocks, most soils are deeply weathered and leached. The red and yellow soils are dominated by kaolinitic clay minerals and aluminium - iron-oxyhydroxides.

Another important feature of the soil is a high infiltration capacity approaching values for sandy soils despite it’s high clay content.

The soil is classified as a "Humic Acrisol" according to the FAO Soil Classification System.

Three soil erosion control measures are tested in 12 classic runoff plots separated into 3 blocks. Each plot with dimensions of 4 times 20 m is fenced on 3 sides by metal plates buried to a depth of 0,5 m. At the downslope end of each plot a collection gutter guides water/soil lost due to surface runoff into a system of 2 concrete tanks. One fourth of an eventual overflow from the first tank is collected in the second tank. The total tank capacity is close to 3000 l. The volume of water and soil lost from the runoff plots is recorded after each single runoff event. Samples are taken for laboratory analyses.

At the site rainfall is recorded by means of an automatic rain gauge monitoring amount and intensity of the incoming rain. The other climatic data is supplied by a nearby meteorological station. Soil water content at different soil depths is monitored the Time Domain Reflectometry (TDR) - method. Soil water samples are collected from 1 m depth in all plots and analysed.

In each of the 3 blocks 3 soil conservation measures are tested against a control plot. The 3 measures are:

- Vetiver grass lines on the contour

- Calliandra hedgerows on the contour

- Tied contour ridges

The vegetative control measures used are Vetiveria zizanioides and Calliandra calothyrsus provided by the HIMA-Iringa project. Grass lines and hedgerows are repeated for every 2 m of vertical interval. Vetiver grass was planted with a spacing of 10 cm between slips, Calliandra with 20 cm between seedlings. Both Vetiver and Calliandra was planted in 1992 and fully effective with regard to soil erosion control in the first trial season 1995/96.

Tied ridges have a spacing of 80 cm.

The control treatment is flat cultivation as practised by nearly all farmers in the area.

In all plots (including the control plots) maize intercropped with beans was used as test crop for all 3 seasons.

All plots received N - and P - fertiliser and pesticides were used when needed.

The first 2 seasons were characterised by a late start of the rain season, rainfall below average, prolonged dry spells and few high intensity rain events. Widely blamed on the "El Nino" phenomenon, rainfall above average with often destructive downpours were recorded in the third season.

In each of the first 2 seasons only one major runoff event was recorded. In total 14 runoff events were recorded in the 1995/96 and 1996/97 seasons and 22 events in the 1997/98 season.

Only 1 - 2% of the incoming rain was lost as surface runoff from control plots in the first 2 season but about 10 % was lost in the last. In single events up to 20 % of incoming rain was lost by surface runoff from control plots during the first 2 seasons and up to 70 % during the last season.

The highest recorded rain intensity was 192 mm/h (5-min. max.) i.e. receiving 16 mm within a period of 5 minutes. No rain storm with an intensity lower than 24 mm/h (5-min. max.) triggered losses of measurable amounts of soil/water. During the first 2 seasons 5-min.-max. intensities had to exceed 40-60 mm/h to cause any runoff since the top soil usually dried up between major rain storms and regained it’s porosity. As the top soil largely was saturated throughout the first months of the last growing season, large amounts of soil and water was lost in rain storms at much lower intensities.

Surface loss of soil and water

All applied soil erosion control measures greatly reduced soil and water loss by surface runoff. Vetiver grass lines virtually stopped the loss of soil and reduced the amount of water lost by surface runoff by 85% on average. In total - 110 tons soil/ha was lost on average from control plots during 3 growing seasons compared to less than 2 tons/ha from Vetiver plots (preliminary data).

Within the 3-year trial period a 50 cm terrace developed along Vetiver grass lines. This is partly due to sediment trapped behind the lines but also due to a downslope cultivation of the plots. The contribution of the latter to a terrace build-up becomes obvious in the "Calliandra plots". Here terraces reached almost the same height although the amount of soil trapped behind the hedges during runoff events is much less.

The efficient filtering effect by Vetiver grass lines is easily observed when comparing soil/water loss in plots with Vetiver lines and Calliandra hedges. Whereas the decreased slope angle/length between the Calliandra hedges sharply reduced surface runoff in the first 2 seasons, an additional filtering effect was needed in the extreme 1997/98 season. During this season average soil loss from Calliandra plots was above 20 tons/ha.

Soil loss from plots with tied ridges was less than 10 tons/ha during 3 growing seasons. When evaluating the effectiveness of tied ridging it has to be mentioned, that ridges were re-established in connection with major agronomic activities in the plots (weeding).

On this soil even a moderate loss of plant nutrients will exceed quantities provided by continuing weathering. Obviously, losses of topsoil from the control treatment, i.e. the commonly practised cultivation of farmland in the area, and its constituents do reach unacceptable values. Although not readily available for plant growth both total N and total P lost from these plots equals the amounts of N and P removed by 6 - 7 tons of harvested maize grain i.e. about 2 years produce for this site (tons/ha) and the applied fertiliser/pesticide strategy.

Plant nutrients lost in this way from Vetiver plots is negligible.

Assuming that most of the water retained in Vetiver plots during runoff events compared to control plots will have to be added to water leaching through the system (rooting zone), then this "surplus percolation" in Vetiver plots will remove more plant nutrients like Ca, Mg, K than are lost by surface erosion. This assumption is based on analyses of soil water samples collected at the bottom of the rooting zone.

As mentioned above the soil water content at different depth in the rooting zone (0-100 cm) is monitored by means of the TDR-technique. Analyses of 3 years of data is still incomplete but with regard to the Vetiver treatment observations do indicate:

a lower soil water content in all layers of the rooting zone inside and close to the Vetiver lines at the end of the dry season as compared to measurements made between the lines and inside the control plots. This may be due to continuously transpirating Vetiver grass throughout the dry season.

a higher soil water content along the Vetiver lines in the upper soil layer in very wet periods of the rainy season (e.g. in Figure 1) as compared to measurements in the control plots. The higher water content may be due to increased infiltration around Vetiver lines during runoff events. A comparable higher water content in the upper horizon is also observed in periods without any surface runoff. This could be explained by a higher water retention capacity of soil accumulated here.

Accumulated within a certain period an increase in soil water content along Vetiver grass lines often exceeds the amount of rain received. This can be due to subsurface flow accumulating at a hydraulic barrier. As seen from Figure 1, this excess water is rapidly lost to drainage/ evapotranspiration in dry spells during the rain season ( day 85 - 105).

a lower soil water content in deeper parts of the rooting zone between Vetiver grass lines i.e. inside the maize/bean crop. A possible explanation to this finding could be a diversion of lateral soil water movement into deeper soil layers along Vetiver grass lines.

Crop yield

Amounts of harvested maize and beans are recorded for each plot. A direct comparison between treatments proved to be difficult since crop yield is affected by changing soil physical/chemical properties within the experimental site. Data on crop yield will have to be collected for some more years before significant differences in crop performance due to soil conservation measures should be considered.

Other findings with focus on Vetiver grass lines

Cultivation at a 29 % slope will leave areas close to the downslope side of Vetiver grass line stripped for humus rich top soil. Crop performance of the first line of maize below Vetiver grass lines therefore was very poor. Along with the area occupied by Vetiver grass this fact was often pointed out by visiting farmers. For farmers with access to farmyard manure and/or compost this problem could be addressed by localised application of these inputs. On the other hand this drastic removal of topsoil accompanied with a poor crop performance can be seen at as a forecast on crop production in unprotected fields.

Rats and mice do find shelter and hiding places in Vetiver grass lines. Although no immediate problems arose from this, a large percentage of maize seed close to the lines was eaten and re-planting had to be done.