Gully Erosion and Stabilization in Semi-arid Environment of Wanjoga River Catchment of Tana Basin, Embu County, Kenya

Abstract

Gullies are prone to occur in semi-arid regions characterized by rainfall variability and increased overland flow, thus affecting thresholds of geomorphic processes and ecological fragility and affecting any landscape restoration required after degradation. Although gully erosion affects less than 5% of the world, most soil loss from gullies generates up to 95% of the global sediment load. As a semi-arid region, gully erosion is the most severe environmental problem in the Wanjoga river catchment of the Tana River basin, Embu County, located about 170 kilometers from Nairobi. The present study examines gully erosion and rehabilitation in a semi-arid environment of Wanjoga river catchment, specifically; (a) evaluating geomorphological factors that initiate and promote progressive development of gully erosion, (b) establishing the relationship between gully morphometry and rate of gully development, c) determine the threshold factors of gully development and (d) to evaluate the suitability of different gully rehabilitation methods used for controlling gully erosion in the semi-arid environment. Data for creating an inventory map of gullied areas, rate of gully development and threshold analysis were obtained by carrying out extensive field surveys and acquiring Landsat images. A series of mapping using Landsat images at 1.5m spot and 15m resolution images were employed to identify the most severely gullied locations. A total of 66 gullied sites in the Wanjoga catchment were mapped. Bivariate statistical analysis to evaluate the influence of conditioning factors on gully development showed a significant positive relationship between gully occurrence and gully conditioning factors with a consistency value at CR =0.097. The positive weighted values on steep slopes > 20° covered by clay lithotypes (Wi=7.53) shows gully erosion occurrence is positively associated with soil lithotype. Weighted Overlay Tool helped categorize gully susceptibility into high (12.73%), moderate (36.32%), and low susceptible zones (46.95%), with steep slopes and soil lithotype playing an imperative role in gully susceptibility at 43% and 25%, respectively. Kappa statistic for the gully susceptibility map confirmed accuracy at approximately 0.42, representing a moderate level of agreement with a positive value of 4% and a false-negative value of 7%. The results suggest the relationship between slope and drainage area in Wanjoga River catchment is S=0.0384-0.397, with 𝑅2 =0.0321 for gully cut area in the upper segment, S=0.174A-0.032, with 𝑅2 =0.498 for gully cut area in mid-segment and S=0.23A-0.020, xv

with 𝑅2=0.088, for gully cut area for the lower segment, representing approximate regions most susceptible for gully initiation. The negative exponent ‘b’ values (⁻0.397 at the upper segment, ⁻0.032 at mid-segment and ⁻0.020 lower segment) revealed a relatively weak regression slope, suggesting a dominance of overland flow erosion. Field survey data was tested for the degree of association between variables using ANOVA (Analysis of Variance), Spearman's rank correlation coefficient and paired t-test to determine the relationship between the variables. For the effectiveness of gully rehabilitation, the t-test reveals a p-value = 0.000, which is less than 0.05, showing that structures for gully rehabilitation used by farmers have not healed a significant number of gullies across all segment regions. Thus, the study suggested; that the reduction of overland flow discharge, vegetative measures, designing structures specified for knick-points and controlling gully head cut erosion are essential procedures in protecting gully erosion in a semi-arid region of varied soil lithotypes. The accuracy of predicting gullied areas and areas more susceptible to gully erosion using the bivariate statistical methodology provides the basis for determining the threshold for gully development which dictates more frequent and elaborate methods of designing and siting conservation structures which provide clues to improve conservation and rehabilitation of semi-arid environments of similar sensitivity to overland flow.