Effect of Soiltac on Wash Erosion by Overland Flow of Some Trinidadian Soils (TPD1012018)

Three soils, Piarco sandy loam, Maracas clay loam and Talparo clay (Table 1) were used for the study. Air-dried soil samples were ground to pass a 5 mm sieve. A particle size analysis (Table 1) was carried out using the hydrometer method (Lambe, 1951). The organic matter content in the samples was measured using the method of Walkley and Black (1934).

Wash erosion was measured using a soil erosion assess- ment facility (Fig. 1) similar to that used by Ekwue et al. (2009) and Ekwue and Harrilal (2010). A water trough was supplied by an adjustable water supply. At the head of the frame holding the apparatus (side with the overflow trough), a winch/pulley system was set up to facilitate the adjustment of the slope of the soil tray. This trough was connected to a rectangular channel section (190 cm long, 28 cm wide and 18 cm high) which acted as a soil tray. A wooden soil frame (190 cm long, 27 cm wide and 4 cm deep), whose bottom was covered with two layers of wire mesh to allow for drainage of water, was made to hold the soil samples during testing. Three pairs of wooden  legs,  each  14  cm  high,  were  placed  on  the  soil  tray such that, when the soil frame was placed on top of the legs, the elevation of the soil frame was aligned with the top of the soil tray. The opposite end of the soil tray was equipped with a 19 cm long, 28 cm wide and 5 cm deep collection tray with a hole at the centre to channel soil sediment and runoff water into a graduated container for measurement. From the runoff, a random sample of 4 l was taken after stirring the soil and water mixture to agitate the soil particles. The water sample was used to determine the soil loss using the standard total soluble solid tests (American Public Health Association, 2005).

To prepare for testing, the wooden soil frame was removed from the soil tray and packed with soil and compacted using a 6 kg, 100 mm diameter steel roller using different passes to achieve the bulk densities of 1.45, 1.42 and 1.20 t m-3  for the sandy loam, clay loam and the clay respectively. These were similar  to  the  existing  densities  of  these  soils  in  the  field. Concentrations of Soiltac recommended by the manufacturer for controlling soil erosion in different slopes (Soiltac website, 2008) were used (Table 2). The manufacturer specified concen- trations required for erosion control in steep, average and light slopes. A 9% slope was considered light, 15% was average slope, 21%  was  average  to  steep  and  30%  was  considered  steep following  the  slope  classification  system  of  Sheng  (1972).  In addition, three other concentrations were used: Control (0%), and 2% by volume less and more than the recommended rates, in order to provide more values for comparison of results. In each case, a 2 l volume of Soiltac with the required concentra- tion was produced and sprayed on to the soils in the soil frame, and  dried  in  the  dehumidification  chamber  for  24  h.  After drying, the wooden soil frame was replaced in the soil tray and the  sides  between  the  tray  and  the  frame  was  sealed  using silicone.

Wash erosion by overland flow was assessed using a facto- rial experiment involving the three soils with the four Soiltac concentrations, and exposed to four slopes (9%, 15%, 21%, and 30%) with two replications. Although the measured wash erosion in this study does not involve erosion by raindrops, measurements of wash erosion with and without raindrops by Ekwue et al. (2009) and Ekwue and Harrilal (2010) on the same three soils used in this test, showed that relative erosion values were maintained in the two tests. Two replications rather than three were adopted because of the consistency of the measurements   and   to   be   within   the   constraints   of the experimental programme. A constant water overland flow rate of  0.25  l  s-1   was  maintained  for  20  min,  but  measurements were made after durations of 5, 10, 15 and 20 min. The 0.25 l s-1 flow  rate represented the lowest rate utilised by  Zhang, Liu, Nearing,   Huang,   and   Zhang   (2002)   and   was   sufficient   to produce measurable values of erosion. It is, however, on the low side of the overland flow rates required to generate gully erosion  in  the  field,  as  recently  Wells,  Alonso,  and  Benneth (2009) used a range of 0.75 l s-1e1.18 l s-1  to achieve this. The slope gradients were chosen to represent the ones prevalent in agricultural soils in Trinidad (Gumbs, 1987). Analysis of vari- ance (ANOVA) of wash erosion values was performed using the MINITAB  Statistical  Software  Release  13.20  by  Minitab  Inc., USA. ANOVA was performed separately for each slope because

Fig. 1 e The soil erosion apparatus.

Soiltac concentrations were not repeated for each slope gradient, since it was considered wiser to test each slope using the values within ±2% by volume of the concentration levels suggested by the manufacturers. Also, in order to avoid the problem of time data being a continuance of the same test replicates, runoff and wash erosion rates rather than the cumulative values were analysed. These rates were obtained by dividing the runoff and wash erosion collected at five minute intervals by five in each case.

3.                Results and discussion

3.1.           Factors affecting wash erosion

Table 3 shows the values of wash erosion for the three soils. Wash erosion decreased with increasing concentration of Soiltac for all combinations of soil type, slope and overland flow duration. Table 4 shows that cumulative runoff increased with increasing slope, and Soiltac concentration. The clay soil maintained slightly greater surface runoff than the clay loam and the sandy soil. As expected, in all cases, wash erosion and cumulative runoff increased with increasing duration of overland flow.

Table 5 summarises the mean rates of runoff and wash erosion for the different experimental factors. Mean values of these parameters were determined for each slope (Table 5) since the concentrations of Soiltac used were different for each slope as shown in Table 2. Mean runoff rates for each slope increased with increasing clay content, Soiltac concentration and dura- tion of overland flow. Generally, mean runoff rates increased with increases in slope. Mean wash erosion rates varied from 0.027 kg m-2 min-1 in the sandy soil to 0.022 kg m-2 min-1 in the clay  loam  soil  for  9%  slope.  At  30%  slope  the  corresponding values were 0.061 and 0.055 kg m-2  min-1. Mean wash erosion dramatically   decreased   with   Soiltac   concentration   at   9% slope   (from   0.085   kg   m-2     min-1     in   the   control   soil   to 0.001 kg m-2 min-1 in the soils with the highest concentration of 19%  by  volume).  At  the  30%  slope,  the  corresponding  values decreased from 0.206 for the control soil to 0.001 kg m-2  min-1 for the soil with a Soiltac concentration of 19% by volume. The mean wash erosion rates for each slope decreased with increasing duration of overland flow.

The analysis of variance (Table 6) showed that the main effects of soil type, Soiltac concentration and overland flow duration were significant in most cases for all four  slopes (P ¼ 0.01) for the runoff and wash erosion rates. The only exceptions were the main effect of soil type for wash erosion at 30% slope and runoff rates at 9% slope, which were not significant. Soiltac concentration was the most important factor that affected wash erosion rates for all the slope gradients. Runoff rates were mostly affected by the duration of overland flow. In addition, the most significant interaction

for the runoff and wash erosion rates was that between Soiltac concentration and duration of overland flow followed by that between Soiltac concentration and soil type for wash erosion rates alone. These were all significant (P ¼ 0.01) and these interactions together with the main effects will be described below. The interaction between soil type and duration of overland flow was also significant in most slopes for runoff rates (Table 6).

3.1.1.        Soiltac concentration

Soil loss rates decreased dramatically (about 99%) with increasing levels of Soiltac concentration in the three soils. This was true irrespective of the slope gradients, and durations of overland flow for all three soils. Previous studies by Ben-Hur, Faris, Malik, and Letey (1989), Levy et al. (1992) and Shainberg and Levy (1994) revealed that PAM also decreased soil loss. Levy et al. (1992) found that PAM in irrigation water reduced sediment loss by up to 98% while Lentz and Sojka (1994) noted a 93% reduction in soil loss. The flow of irrigation water is expected to be similar to the overland flow used in this research since raindrop energy was not involved. The present study therefore shows that Soiltac is as effective as the other polymers in the sense that once it is applied, it more or less eliminates soil loss. However, unlike PAM used in previous studies, present results showed that the incorporation of Soiltac increased runoff from the soil. Shainberg et al. (1990) and Smith et al. (1990) observed that addition of PAM resul- ted in increased infiltration rates by 7e8 fold compared to the control, meaning lower runoff rates. Shainberg and Levy (1994) on the other hand noted a 10 fold increase. In the present studies, Soiltac applied at the recommended rates increased runoff rates by about 2.2% for the 30% slope to 6.4% in the 9% slope. This means lower infiltration into the soil. This is in contrast with previous research by Ekwue and Harrilal (2010) which found that soil amendments like peat reduced soil loss by increasing soil infiltration and decreasing runoff rates. Previous studies on PAM showed that it reduced soil loss by stabilising the soil, which helps it to maintain adequate pore

Fig. 2 e Effect of the interaction between Soiltac concentration levels and soil type on wash erosion rates (Data for 9% slope were plotted).

space for infiltration (Green & Stott, 2001) thereby reducing runoff. In the present study, it was found that, on application of Soiltac and eventual drying of the soil, it became very hard and this reduced infiltration and increased runoff rates. Soil strength is known to increase the resistance of soils to erosion (Rachman, Anderson, Gantzer, & Thompson, 2003; Wuddivira, 2008). The implication of this study is that the use of Soiltac in agricultural fields at the recommended concentrations is undesirable, since the aim is to allow water infiltration into the soil in order to encourage crop growth. Increased runoff in agricultural fields is not only problematic to soil and water conservation, but also increases the risk of nutrient discharge (Bjorneberg, Aase, & Westermann, 2000). However, as was also found for PAM, preliminary field tests conducted by the present authors showed that Soiltac could be very effective in reducing dust in dirt road construction. The manufacturers of Soiltac recommended its use in soil slope stabilisation works. This study has shown that adequate drainage facilities should be  provided  in  these  works  to  take  care  of  increased  runoff generated. The manufacturers of Soiltac also recommended its use in sealing the sides of runoff ponds and the reduction of infiltration  observed  in  this  study  supports  this  assertion.  A comparative study of polymers will be conducted in future to find  out  why  Soiltac  acted  the  way  it  did.  Another  major difference observed for the PAM and Soiltac is the application rates. For instance, Shainberg et al. (1990) used the normally recommended rates of 20 kg ha-1 of PAM which is much lower than the application rate of about 4510 kg ha-1  recommended by  the  manufacturers  of  Soiltac  for  soil  erosion  control  on a steep slope.

The significant interaction between Soiltac concentration level and soil type (Fig. 2) shows that although differences in wash erosion rates were significant for the three control soils with no Soiltac, once Soiltac was applied, the differences in these parameters were minimal. The interaction between Soiltac concentration level and overland flow duration (Fig. 3) implies that although wash erosion rates decreased and runoff rates increased with increasing duration of overland flow for the control soil, when Soiltac was applied, wash erosion rates dropped dramatically and runoff rates increased significantly, but the effect of greater concentration of Soiltac was minimal. All these interactions reinforce the fact that although surface application of Soiltac at the recommended rates will dramatically affect soil properties related to soil erosion, values of these properties will remain constant with increased  Soiltac concentration.

3.1.2.        Soil type

The main effect of soil type on wash erosion rates and runoff rates was significant (Table 6). As determined in the previous studies by Ekwue and Harrilal (2010), soil type was the least important of all the three experimental factors (Table 6). Piarco sandy loam had the largest quantity of mean soil loss (Tables 3 and 5). Although this soil had the lowest values of surface runoff (Tables 4 and 5), as was found in previous studies by Ekwue and Harrilal (2010), its low percentage clay content  (18.1%,  Table  1)  decreased  the  soil  strength  thus decreasing the soil’s ability to increase the cohesiveness of the particles and pose resistance for the overflow flow. However, in the present study, the interaction obtained between soil type and Soiltac concentration levels (Fig. 2) means that these differences in wash erosion will be minimal once Soiltac is applied to stabilise the soils.

3.1.1.        Slope

Concentrations of Soiltac applied for different slopes varied and so analysis of variance was done separately for each slope. However, values in Tables 3 and 5 show that, as expected, soil loss increased significantly in each case with slope gradient. At a higher slope gradient, there was an increase in the velocity of water over the surface. The increase in water velocity resulted in a greater erosive power of the water. The measured values (Tables 4 and 5) showed that wash erosion also increased with slope gradient as a result of greater runoff in steeper slopes when compared to gentler slopes. Data in Table 5 showed that the effectiveness of Soiltac in increasing runoff rates decreases with increasing slope. However, for wash erosion, data in Table 5 showed that the reverse was the case because the maximum effect of Soiltac was obtained in steep rather than gentler slopes.

3.1.4.     Overland flow duration

As expected, mean cumulative runoff as well as wash erosion increased with increasing overland flow duration in all the soils, no matter the concentration of Soiltac and soil slope. This is because of the increased volumes of overland flow obtained as the duration increased. However, while the wash erosion rates declined, the runoff rates increased with increasing duration of overland flow (Table 5).

Epstein and Grant (1967), Jennings, Jarrett, and Hoover (1987) and Ekwue and Ohu (1990) have all found that soil detachment rates decreased with increasing rainfall duration and attributed this to the compaction of the surface soil by raindrops and to the breakdown of large aggregates during rainfall. In the present study, the reduction of wash erosion rates with time by overland flow could be related to the depletion of detachable soil particles as the duration of over- land flow proceeded. Soil erosion became source limited. On the other hand, the increased rates of runoff with the duration of overland flow could be attributed to increased soil sealing and decreased infiltration rates as the overland flow pro- gressed. This will be further investigated in a future study. The concentration (Fig. 3) implies that, when Soiltac is applied to the soil, the variations in runoff and wash erosion rates with overland flow duration are minimal.

3.2       Derivation of regression equation relating wash erosion to experimental factors

For each slope gradient, the wash erosion rates for the three soils with four Soiltac concentration levels, and four overland flow durations were used to generate multiple linear regres- sion equations that could be used to predict wash erosion rates. The equations were of the  form:

C is clay interaction   between   overland   flow   duration   and Soiltac where  Ew   is  wash  erosion  rates  (kg  m-                 –             t content of the soil (%), Conc is Soiltac concentration (%), Q is the cumulative volume of overland flow (m3) and a, b, c, and d are the empirically derived coefficients (Table 7) for each of the four slopes used in the study. The signs of the experimental factors obtained confirm how the factors affected the wash erosion rates. The multiple correlation coefficients (Table 7) for each slope were significant at 0.1% level. The Student’s ‘t’ values for all the experimental factors (Table 7) except those for clay content were significant at 0.1% level. The relative ‘t’ values for all the factors also confirm the findings in the analysis of variance that the most important factors that affected wash erosion rates for all slopes were Soiltac concentration followed by the duration of overland flow. Clay content was the least important factor and the Student’s ‘t’ values for clay were found not to be significant for all the soil slopes.

4.                Conclusions

Wash erosion by overland flow was measured for three Trinidadian soils in the laboratory using a specially constructed soil erosion apparatus. Soiltac was very effective in almost eliminating wash erosion in the soils. Wash erosion was greatest in Piarco sandy loam and was smallest in the clay loam soil. Soiltac increased runoff volumes, meaning that its use to reduce soil erosion in agricultural soils may be undesirable except in very special cases. However, it may be useful in construction sites and dirt road works, although there should be provision of drainage works to take care of increased runoff. Soil loss increased with increasing soil slopes and overland flow duration, although the increases were minimal for soils with Soiltac treatment. The interaction effects obtained in this study showed that, much as the surface application of Soiltac is expected to increase runoff and reduce soil loss, the concentration levels suggested by the manufacturers could be varied by ± 2% by volume without greatly affecting the results. Nothing will be gained by slightly decreasing or increasing the recommended rates. Multiple regression equations derived to relate wash erosion rates for different slopes to the experimental factors were highly significant and confirmed that the most important factors that affected soil loss rates were Soiltac concentration and duration of overland flow. The implication of this study is that, while land use zoning of soils based on slopes is essential in soil conservation, the incorporation of Soiltac will minimise soil erosion by water. It will, however, also decrease infiltration and increase runoff, making it undesirable in agricultural fields.

references

Ahmad, N., & Breckner, E. (1974). Soil erosion on three Tobago soils. Tropical  Agriculture,  51, 313e324.

American Public Health Association (2005). Total suspended solids dried at 103e105o C. Part 2540-D. In A.D. Eaton, L.S. Clescert, E.W. Rice, A.E. Greenburg, & M.A.H. Franson, (Eds.), Standard methods for the examination of water and Wastewater, 21st ed., Washington D.C. 2-58 to  2-59.

Ben-Hur, M., Faris, J., Malik, M., & Letey, J. (1989). Polymers as soil conditioners under consecutive irrigations and rainfall. SoilScience Society of America Journal, 53,  1173e1177.

Bjorneberg, D. L., Aase, J. K., & Westermann, D. K. (2000). Controlling sprinkler irrigation runoff, erosion and phosphorus loss with straw and polyacrylamide. Transactionsof the  ASAE,  43, 1545e1551.

Ekwue, E. I., Bharat, C., & Samaroo, K. (2009). Effect of soil type, peat and farmyard manure addition, slope and their interactions on wash erosion by overland flow of some Trinidadian soils. Biosystems Engineering, 102,  236e243.

Ekwue, E. I., & Harrilal, A. (2010). Effect of soil type, peat, slope, compaction effort and their interactions on infiltration, runoff and raindrop erosion of some Trinidadian soils. BiosystemsEngineering, 105, 112e118.

Ekwue, E. I., & Ohu, J. O. (1990). A model equation to describe soil detachment by rainfall. Soil  and  Tillage  Research,    16, 299e306.

Epstein, E., & Grant, W. J. (1967). Soil losses and crust formation as related to some soil physical properties. Soil Science Society ofAmerica Proceeding, 31, 547e550.

Green, V. S., & Stott, D. E. (2001). Polyacrylamide: a review of the use, effectiveness, and cost of a soil erosion control amendment. In D. E. Stott, R. H. Mohtar, & G. C. Steinhardt (Eds.), Sustaining the Global Farm, 10th International soil conservationorganization meeting (pp. 384e389). Purdue University and USDA-ARS National Soil Erosion Research Laboratory.

Gumbs, F. A. (1987). Soil and water conservation methods for theCaribbean, Trinidad. Department of Agricultural Extension, University of the West Indies.

Jennings, G. D., Jarrett, A. R., & Hoover, J. R. (1987). Simulated rainfall duration and sequencing affect soil loss. Transactionsof the American Society of American Engineers, 30, 158e161,   165.

Lambe, T. W. (1951). Soil testing for Engineers. New York: John Wiley.

Lentz, R. D., Shainberg, I., Sojka, R. E., & Carter, D. L. (1992). Preventing irrigation furrow erosion with small applications of polymers. Soil Science Society of America Journal, 56,  1926e1932.

Lentz, R. D., & Sojka, R. E. (1994). Field results using polyacrylamide to manage furrow erosion and infiltration. SoilScience, 158, 274e282.

Levin, J., Ben-Hur, M., Gal, M., & Levy, G. J. (1991). Rain-energy and soil amendment effects on infiltration and erosion of three different soil types. Australian Journal of Soil Research, 29, 455e465.

Levy, G. J., Levin, J., Gal, M., Ben-Hur, M., & Shainberg, I. (1992). ’Polymers’ effects on infiltration and soil erosion during consecutive simulated sprinkler  irrigations.  Soil  Science  Societyof  America  Journal,  56,  902e907.

Mahabir, R., & Al-Tahir, R. (2008). The role of spatial data infrastructure in the management of land degradation in small tropical Caribbean Islands. In: Tenth InternationalConference for Spatial Data Infrastructure, 25e29 February, St. Augustine, Trinidad.

Morgan, R. P. C. (2005). Soil erosion and conservation (3rd ed.). New York:  John Wiley.

Orts, W. J., Sojka, R. E., & Glenn, G. M. (2000). Biopolymer additives to reduce soil erosion-induced soil losses during irrigation. Industrial Crops  and  Products, 11, 19e29.

Quansah, C. (1981). The effect of soil type, slope, rain intensity and their interactions on splash detachment and transport. Journal of Soil Science, 32,  215e224.

Rachman, A., Anderson, S. H., Gantzer, C. J., & Thompson, A. L. (2003). Influence of longterm cropping systems on soil physical properties related to soil erodibility. Soil Science Societyof America Journal, 67,  637e644.

Shainberg, I., & Levy, G. J. (1994). Organic polymers and soil sealing in cultivated soils. Soil Science, 158,   267e273.

Shainberg, I., Warrington, D. N., & Rengasamy, P. (1990). Water quality and PAM interactions in reducing surface sealing. SoilScience, 149, 301e307.

Sheng, T. C. (1972). ATreatment-orientedlandcapabilityclassification SchemeforHillyMarginallandsintheHumidTropics. FAO Report No. TA 3112, Rome.

Smith, H. J. C., Levy, G. J., & Shainberg, I. (1990). Water-droplet energy and soil amendments: effect on infiltration and erosion. Soil Science Society America Journal, 54,   1084e1087.

Soiltac. (2008). Powdered soiltac product information for soil stabilization. http://www.powderedsoiltac.com Accessed 31.07 10.

Soil Survey Staff. (1999). Soil Taxonomy: a basic system for making and interpreting soil surveys. In (second ed). Agriculture Handbook, vol. 436 (pp. 128e129) Washington, DC: USDA, US Govt Printing Office.

Walkley, A., & Black, I. A. (1934). An examination of the effect of Degtjareff method for determining soil organic matter  and a proposed modification of the chromic acid titration method. Soil Science, 37,  29e38.

Wells, R. R., Alonso, C. A., & Benneth, S. J. (2009). Morphodynamics of headcut development and soil erosion in upland concentrated flows. Soil Science Society of AmericaJournal,  73, 521e530.

Wuddivira, M. N. (2008). Structural stability, hydraulic properties anderodibility of humid tropical soils under intense rainfall. PhD Thesis. St. Augustine, Trinidad, West Indies: The University of the West Indies.

Wuddivira, M. N., Ekwue, E. I., & Stone, R. J. (2010). Modeling slaking sensitivity to assess the degradation  potential  of humid tropical soils under intense rainfall.  Land Degradationand  Development,  21,  61e73.

Zhang, G. H., Liu, B. Y., Nearing, M. A., Huang, C. H., & Zhang, K. L. (2002). Soil detachment by shallow flow. Transactions  of   theASAE,   45,  351e357.