ABSTRACT
Rainfed farming is a dominant agricultural system in Tigray, Ethiopia. However, rainfall is characterized by short duration, intense and erratic subjected to late-onset and early cessation, suggesting a pressing need for Supplemental Irrigation (SI) to fill the crop water demand. Understanding the effects of SI during rainfall late-onset, early cessation, or both, along with their underlying causes, is a critical knowledge gap globally. Although wheat is one of the principal food crops in Tigray, it is subjected to moisture stress during critical growth stages, limiting its potential productivity. Studies specifically related to impacts due to the late-onset of rainfall on wheat are non-existent. Here, we investigated the agrometeorological characteristics of rainfall variability, onset, cessation, and length of the growing season to evaluate the use of SI for balancing the moisture stress in rainfed farming. Meanwhile, using an on-farm experiment, we also evaluated double-season (2017 and 2018) SI application during late-onset (Pre), early cessation (Post), and its combined effects (Pre + Post) on yield and water productivity (WP) of wheat. Yield and WP were significantly (P < 0.05) affected by SI application with higher grain yield (3298 kg/ha) and WP (0.538 kg/m3) obtained from applying Pre + Post. Applying Pre + Post has increased grain yield, biomass, and WP of wheat by 45.6, 27.7, and 21.5% over Rain-fed farming, respectively. Thus, balancing crop water requirements using SI during inadequate rainfall distribution is key for improving WP and wheat production in semi-arid environments. Particularly, the application of SI both during the late-onset and early cessation of rainfall is suggested for greater wheat productivity in semi-arid regions.
ABSTRACT
Understanding the relationship between hydrological processes and environmental changes is important for improved water management. The Geba catchment in Ethiopia, forming the headwaters of Tekeze-Atbara basin, was known for its severe land degradation before the recent success in integrated watershed management. This study analyses the hydrological response attributed to land management change using an integrated approach composed of (i) simulating the hydrological response of Land Use/Cover (LULC) changes; (ii) assessing the alteration of streamflow using Alteration of Hydrological Indicators (IHA); and (iii) quantifying the contribution of individual LULC types to the hydrology using Partial Least Square Regression model (PLSR). The results show that the expansion of agricultural and grazing land at the expense of natural vegetation has increased the surface runoff 77% and decreased dry season flow by 30% in the 1990s compared to 1970s. However, natural vegetation started to recover from the late 1990s and dry season flows increased by 16%, while surface runoff declined by 19%. More pronounced changes of the streamflow were noticed at sub-catchment level, mainly associated with the uneven spatial distribution of land degradation and rehabilitation. However, the rate of increase of low-flow halted in the 2010s, most probably due to an increase of water withdrawals for irrigation. Fluctuations in hydrological alteration parameters are in agreement with the observed LULC change. The PLSR analysis demonstrates that most LULC types showed a strong association with all hydrological components. These findings demonstrate that changing water conditions are attributed to the observed LULC change dynamics. The combined analysis of rainfall-runoff modelling, alteration indicators and PLSR is able to assess the impact of environmental change on the hydrology of complex catchments. The IHA tool is robust to assess the magnitude of streamflow alterations obtained from the hydrological model while the PLSR method is useful to zoom into which LULC is responsible for this alteration.
ABSTRACT
The spatiotemporal variability of the Land Use/Cover (LULC) is a strong influence on the land management and hydrological processes of a river basin. In particular in semi-arid regions like the Tekeze-Atbara (T-A) basin, accurate information about LULC change is a prerequisite for improved land and water management. The human-induced landscape transformations in the T-A basin, one of the main tributaries of the Nile River, were investigated for the last four decades (1972-2014). Separate LULC maps for the years 1972, 1989, 2001, and 2014 were developed based on satellite images, Geographic Information System (GIS) and ground information. Change detection analysis based on the transitional probability matrix was applied to identify systematic transitions among the LULC categories. The results show that >72% of the landscape has changed its category during the past 43years. LULC in the basin experienced significant shifts from one category to other categories by 61%, 47%, and 45%, in 1972-1989, 1989-2001, and 2001-2014, respectively. Although both net and swap (simultaneous gain and loss of a given LULC during a certain period) change occurred, the latter is more dominant. Natural vegetation cover, including forests, reduced drastically with the rapid expansion of crops, grazing areas and bare lands during the first two decades. However, vegetation started to recover since the 1990s, when some of the agricultural and bare lands have turned into vegetated areas. Forest land showed a continuous decreasing pattern, however, it has increased by 28% in the last period (2001-2014). In contrast, plantation trees have increased by 254% in the last three decades. The increase of vegetation cover is a result of intensive watershed management programs during the last two decades. The driving forces of changes were also discussed and rapid population growth and changing government policies were found to be the most important.
