In the study area located in Western Kenya near the Lake Victoria, severe soil erosion occurred and it thought to relate to vegetation degradation caused by overgrazing. The livestock density estimated by analyzing satellite image (1.39 TLU/ha for available grazing lands) was lower than that of measured for seven farmers’ grazing lands using GPSs (4.41 TLU/ha, 2011) with variation from 0.83 to 12.36 TLU/ha. Thus, it is clear that the grasslands used by farmers are limited compared with the area of estimated available land for grazing identified by analyzing the satellite image. According to growth-consumption rate model that was developed by the Nyangito et al. (2008) in southeastern Kenya, if livestock density reaches over 7 TLU/ha, pasture growth rate became lower than consumption rate. Grass biomasses of the grazing lands were kept low (less than 50 g/50 × 50 cm 2) under high livestock density (three farmers out of seven were higher than 7 TLU/ha). In addition, rainfall pattern is very unstable and we observed stunted growth of grasses during dry spells. Therefore, we concluded that overgrazing. It means that inhibition of continuous re-growth of grasses due to high grazing pressure has been occurred even for small area and contributed to the soil erosion.
It is well known that topsoil is being depleted through various human activities. While topsoil develops over centuries, the world’s growing human population is actively depleting this resource over decades [
Western Kenya has one of the highest rural population densities in Africa, and soil erosion occurs both often and severely [
In our study area, information about population and density of livestock is very limited and descriptive [
The Nyando River runs from the highlands of Nyando Province about 2000 m above sea level (asl) to Lake Victoria at 1184 m asl. Four hundred and fifty to 500 years ago or earlier, the Luo, Nilotic people who emigrated from Sudan to the northern coastline of Lake Victoria, included the Nyando River basin [
Average annual rainfall around Lake Victoria exceeds 1000 mm/year. In fact, according to the rainfall data at the Kibos Center of Kenya Agricultural Research Institute (KARI), located about 50 km northeast of the study area, the average annual rainfall for 57 years, from 1952 to 2008, was 1306 mm/year, the long rains peaking between April and May and the short rains between November and December.
Daily precipitation in the study area was measured by rain gauge (3639 pulse, Hioki Electric LTD., Japan) from April 2, 2009 to March 18, 2010, and is shown in
The author Yamane stayed with a host family in the study area for a total of seven months, from March to May 2009, January to March 2010, and February 2011. The daily life of the people and their agricultural ways, were
determined using the participatory observation method, and the following investigations were conducted.
An RGB image of the study area was created from Digital Globe’s orthorectifed Quickbird image taken on April 14, 2009 using GIS software, GRASS6.4.0. Most of the study area was located in the Luo land in the Jimo East sub-location, but extended partly in to the Kipsigis lands, which were rented by the Luo people (
Levels of vegetation in each cultivated, grassland, and household compartments were classified as level 0, 1, 2, 3, or 4 following the criteria shown in
The location and size of compartments used for grazing were surveyed six times within three years; twice in April and May 2009, three times in February and March 2010, and once in February 2011. This was performed using GPS (SONY PCQ-HGR3S in 2009, Germin e-Trex Venture HC in 2010 and 2011, respectively) and a byte counter [
Farmers keeping a large number of cattle need more grasslands for grazing, compared with those keeping a
Vegetation level* | Household compartment | Cultivated compartment | Grassland compartment | Total | ||||
---|---|---|---|---|---|---|---|---|
Compartment | Average land size | Compartment | Average land size | Compartment | Average land size | Compartment | Average land size | |
(Number) | (ha) | (Number) | (ha) | (Number) | (ha) | (Number) | (ha) | |
Level 0 | -** | - | 355 | 0.27 ± 0.20 | - | - | 355 | 0.27 ± - |
Level 1 | 25*** | 2.5 ± 0.37*** | 465 | 0.33 ± 0.26 | - | - | 490 | 0.44 ± - |
Level 2 | 342 | 0.39 ± 0.15**** | - | - | 534 | 0.61 ± 0.58 | 876 | 0.52 ± 0.49 |
Level 3 | 164 | 0.39 ± 0.16 | - | - | 108 | 0.81 ± 0.63 | 272 | 0.56 ± 0.51 |
Level 4 | 10 | 0.27 ± 0.12 | - | - | 11 | 0.36 ± 0.22 | 21 | 0.32 ± 0.19 |
Total | 541 | 0.48 | 820 | 0.30 | 653 | 0.64 | 2014 | 0.46 |
*: Criteria of the classification of each vegetation level was shown in
small number. In a preliminary participatory observation, two grazing patterns, a) grazing only inside the household compartment, and b) grazing both inside and outside were detected. The difference in these grazing patterns depended on the number of cattle. Farmers keeping only one or two cattle tended to follow pattern a),
Criteria | Green vegetation | Tree |
---|---|---|
Level 0 | None | (None or a few) |
Level 1 | Less than half land area | (None or a few) |
Level 2 | More than half land area | (None or a few) |
Level 3 | More than half land area | Less than half land area |
Level 4 | More than half land area | More than half land area** |
*: A compartment was classified based on the land area of green vegetation and the land area of the presence of trees. **: Commonly, trees were observed within land area that grass was grown.
whereas those keeping more cattle tended to follow pattern b). Thus, farmers following pattern b) might have a bigger impact on reducing land vegetation than those adopting pattern a). Therefore, in this survey, these two types of farmers were chosen.
The survey was carried out over five successive days, by attaching GPS and bite counters to the necks of cattle. Four farmers keeping more than five cattle (Households A, B, C, and D) and three farmers (Households A, C and D) were selected in 2009 and 2010, respectively (Figures 2(a)-(c)). One farmer (Household E) keeping only two cattle was employed for comparison in 2010. In February 2011, five more farmers keeping a large number of cattle (Households G, H, I, J, and K) were employed and their sheep were used in the survey.
While surveying livestock behavior, the biomass of grasses from four compartments (Households A, B, C, and D in 2009, Households A, C, D and E in 2010) was measured. Three grass samples were taken at random from each household compartment for each measurement. Seven measurements were made for Household B at weekly intervals for two months, from April to early June, in 2009. Four measurements were made in the same period in Households A, C, and D. In 2010, three measurements were made in the two months, from February to March. All grasses within a square area of 50 × 50 cm2 were cut at ground level for dry weight analysis. All harvested grasses were dried in paper bags in the oven at 70˚C for three days before weighing. In 2011, the biomass from 10 compartments was measured on August 5 and 6. This included seven grassland and three household compartments, distributed throughout the study area to determine the variation in biomass among different locations.
A rain gauge (3639-20 Pulse Logger, Hioki E. E. Corporation, Japan) was set up on April 2, 2009 to measure daily rainfall in the inner grass garden of Dori-Secondary School located within the study area.
In the study area, many farmers raised domestic animals such as cattle, goats, and sheep on grazing land. Dairy cattle and dairy goats are reared for milk and other small ruminants, goats and sheep, are reared for meat. Households were classified into five types based on the livestock farming composition of animals: Group A (cow + bull + small ruminants), B (cow + small ruminants), C (cow or bull), D (small ruminants), and E (no animals) (
The 541 households were classified in
The compartments were categorized based on their vegetation levels (
Forty five surveyed households | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Group* | Household | Head-count/household** | ||||||||||||
Number | Proportion (%) | Cow | Bull | Small ruminants | ||||||||||
Group A | 7 | 15.6 | 3.9 ± 2.3 | 2.7 ± 0.9 | 4.6 ± 5.3 | |||||||||
Group B | 13 | 28.9 | 2.0 ± 1.0 | 0.0 ± - | 4.2 ± 3.1 | |||||||||
Group C | 5 | 11.1 | 1.8 ± 0.5 | 6.0 ± - | 0.0 ± - | |||||||||
Group D | 13 | 28.9 | 2.6 ± 1.8 | |||||||||||
Group E | 7 | 15.6 | ||||||||||||
Total | 45 | 100.0 | ||||||||||||
Whole 541 households in the study area*** | ||||||||||||||
Group | Household | Number of livestock (estimated) | ||||||||||||
Number (estimated) | Cow | (TLU)**** | Bull | (TLU) | Small ruminants | (TLU) | ||||||||
Group A | 84 | 328.2 | (229.5) | 224.4 | (156.9) | 387.1 | (38.7) | |||||||
Group B | 156 | 312.6 | (218.6) | 0.0 | (0.0) | 656.4 | (65.6) | |||||||
Group C | 60 | 86.6 | (60.5) | 72.1 | (50.4) | 0.0 | (50.4) | |||||||
Group D | 156 | 0.0 | (0.0) | 0.0 | (0.0) | 406.4 | (40.6) | |||||||
Group E | 84 | 0.0 | (0.0) | 0.0 | (0.0) | 0.0 | (50.4) | |||||||
Total | 541 | 727.3 | (508.6) | 296.5 | (207.4) | 1449.9 | (145.0) | |||||||
*: Households were classified into 5 Groups based on the Group of keeping livestock: Group A (cow + bull + small ruminant), Group B (cow + small ruminant), Group C (cow or bull), Group D (small ruminants), Group E (no livestock). **: The figure is shown as mean ± standard deviation. ***: The figures of households or livestock calculated number based on the proportion of households or the mean value of livestock shown in the results of 45 surveyed households, respectively. ****: TLU: Tropical livestock units, one TLU equal to 250 kg live weight and it reported as 1.43 cattle (cow or bull) or 10 small ruminants (D. Bourn and W. Wint, 1994).
classified based on their vegetation levels into level 1 (25, 4.6%), level 2 (342, 63.2%), level 3 (164, 30.3%), and level 4 (10, 1.8%) (
Head-count of cattle (cow and bull) or of small ruminants can be converted into tropical livestock units (TLU) for ease of comparison [
Wint and Bourn (1994) reported that TLU is strongly correlated with mean annual rainfall in sub-Saharan Africa [
Variations were observed in the dry weight of grasses among different household compartments, harvesting times, and between years (
In fact, it was extremely dry in March 2009 (data not shown), and conditions were unfavorable for the growth of grasses, resulting in mean dry weights of less than 10 g per unit area (50 × 50 cm2) (
Even though low grass weights were observed in some areas, the dry weight of grasses varied from 2.5 to 23.7 g (12.3 g average) in 2009, 3.6 to 53.9 g (23.9 g average) in 2010, and 6.7 to 46.3 g (22.9 g average) in 2011, respectively (
Farmers kept their animals in cattle pens (“kul” in Luo language) at night, located in the center of their household compartments. Cattle were released from kul early in the morning, and started grazing around the house. Before, or around noon, farmers took cattle to the grassland compartment where they remained until evening. Bite counter showed that grazing started between 8:00 and 9:00 am whilst in their household compartments (
The movements of cattle for grazing in April and May 2009, February and March 2010, and February 2011 were traced by GPS (
In 2010, the farmer from Household A used three different grassland compartments from those used in 2009,
Livestock number | ||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Number of compartments | Total land size (ha) | cattle | small ruminants | |||||||||||
2009 | 2010 | 2011 | 2009 | 2010 | 2011 | 2010 | 2011 | 2009 | 2010 | 2011 | 2011 | |||
Farmers* | Apr | May | Feb, Mar | Feb | Apr | May | Feb, Mar | Feb | Apr, May | Feb, Mar | Feb | Feb | ||
Household A | 4 | 4 | 4 | 4 | 9.3 | 4.2 | 1.7 | 6.2 | 11 | 10 | 11 | 11 | ||
Household B | 3 | 3 | - | 3 | 5.6 | 5.6 | - | 1.2 | 11 | - | 11 | 44 | ||
Household C | 3 | 1 | 1 | - | 13.7 | 5.0 | 4.8 | - | 12 | 10 | - | - | ||
Household D | 1 | 1 | 2 | - | 2.2 | 2.2 | 2.3 | - | 11 | 7 | - | - | ||
Household E | -** | - | 1 | - | - | - | 0.1 | - | - | 2 | - | - | ||
Household F | - | - | - | 2 | - | - | - | 1.7 | - | - | 10 | 4 | ||
Household G | - | - | - | 3 | - | - | - | 1.2 | - | - | 7 | 9 | ||
Household H | - | - | - | 3 | - | - | - | 4.4 | - | - | 17 | 32 | ||
Household I | - | - | - | 1 | - | - | - | 4.4 | - | - | 5 | 10 | ||
Household J | - | - | - | 1 | - | - | - | 0.4 | - | - | 6 | 6 | ||
*: See
and changed compartments every day (
Seven farmers were surveyed in 2011, three in the northern part and four in the southern part of the study area (
Land size of grazing lands, livestock numbers, livestock density and grass density of 10 households are shown in
In the case of Household C, land sizes used for grazing in May drastically reduced from April in 2009, whilst the livestock density increased (0.61 TLU/ha in April to 1.66 TLU/ha in May) (
Livestock density (TLU/ha) | Available grasses for animals (kg/TLU)** | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2009 | 2010 | 2011 | 2009 | 2010 | 2011*** | ||||||||
cattle | cattle | cattle | Cattle + small ruminant | cattle | cattle | cattle | Cattle + small ruminant | ||||||
April | May | Feb-Mar | February | April | May | February | March | February | |||||
Farmers | (2nd to 6th) | (20th to 24th) | (20th to 24th) | (22th to 26th) | |||||||||
Household A | 0.83 | 1.82 | 4.09 | 1.25 | 1.34 | 521.9*** | 285.4 | 221.1 | 181.9 | 312.0 | 596.9 | 555.1 | |
Household B | 1.37 | 1.37 | - | 6.44 | 8.37 | 314.9 | 379.1 | - | - | 0.0 | 115.6 | 88.9 | |
Household C | 0.61 | 1.66 | 1.44 | - | - | 703.2 | 312.6 | 626.1 | 515.3 | 883.8 | |||
Household D | 3.51 | 3.51 | 2.14 | - | - | 123.1 | 148.1 | 423.3 | 348.4 | 597.5 | |||
Household E | -**** | - | 9.54 | - | - | - | - | 94.7 | 78.0 | 133.7 | |||
Household F | - | - | - | 4.15 | 4.27 | - | - | - | - | - | 179.4 | 174.2 | |
Household G | - | - | - | 4.03 | 4.43 | - | - | - | - | - | 184.4 | 168.1 | |
Household H | - | - | - | 2.69 | 3.08 | - | - | - | - | - | 276.1 | 241.9 | |
Household I | - | - | - | 0.80 | 0.92 | - | - | - | - | - | 929.0 | 807.5 | |
Household J | - | - | - | 11.50 | 12.36 | - | - | - | - | - | 64.7 | 60.2 | |
*: Livestock density was calculated based on the data shown in
Household C in April 2009 (13.7 ha) at the beginning of the rainy season was about three times larger than that in May 2009 (5.0 ha) and January to March 2010 (4.8 ha). Households A and C changed grassland compartments in April and May (
Within the study area, farmers can rent land for grazing if required. However, the cost of land rental is high. For example, farmer D paid 1700 Kenyan shillings (Ksh) (1 Ksh = 0.011 USD) to rent a small piece of land for one year (0.35 ha) in 2010, which is approximately twice the cattle herder’s monthly income of 800 Ksh. These high prices force most farmers in this area to use the same lands to raise their animals, although the grass condition or biomass of grasses is limited for grazing.
Nine farmers adopted pattern b) grazing, and one farmer, from Household E, adopted pattern a) grazing (
Small ruminants were usually taken to the same grassland compartments as cattle. In 2011, the size of the grazing land of Household A was 6.2 hectares, whereas Household J used 0.4 hectares, approximately 1/15 in size of that used by Household A (
The amount of grass per head of cattle in 2011 was estimated to be 596.9 kg/TLU for Household A and 64.7 kg/TLU for Household J (
According to the 57 years of rainfall data measured at the Kenya Agricultural Research Institute (KARI) in Kibos, in this region rain often stops for more than two weeks, once or twice in a year. Furthermore, every few years, rain may stop for a whole one month (personal communication). The repeating periods of heavy rain and dry spells are a common characteristic of the rainfall pattern in many tropical regions. The heavy rain and dry spell is likely to be a major factor causing soil erosion [
The average livestock density of seven farmers in February 2011 was 4.41 TLU/ha (
Because of a lot of rain (1546.5 mm/year,
In addition, based on the growth-consumption rate model developed for the agro pastoral system of southeastern Kenya with the annual rainfall 1200 mm in the highlands, at higher stocking rates (above 7 TLU/ha) pasture growth rate initially lags behind consumption [
YukoYamane,ShuichiAsanuma,KazuhiroUmenura, (2015) Influence of Livestock Farming on Vegetation in a Degraded Soil Area on the East Coast of Lake Victoria in Western Kenya: A Case Study of Jimo East Sub-Location in Nyando Sub-County. Journal of Environmental Protection,06,824-836. doi: 10.4236/jep.2015.68075